Bioavailability of Cadmium, Copper, Nickel and Zinc in Soils Treated with Biosolids and Metal Salts
A thesis submitted in partial fulfilment of the requirements for the Degree of Doctor of Philosophy at Lincoln University by Amanda Black
Lincoln University 2010
Abstract of a thesis submitted in partial fulfilment of the requirements for the Degree of Doctor of Philosophy. Abstract
Bioavailability of Cadmium, Copper, Nickel and Zinc in Soils Treated with Biosolids and Metal Salts
by Amanda Black It is widely accepted that bioavailability, rather than total soil concentration, is preferred when assessing the risk associated with metal contamination. Despite this, debate continues on what constitutes a bioavailable pool and how to best predict bioavailability, especially in relation to crop plants. The overall aim of this thesis was to assess and validate measures of cadmium (Cd), copper (Cu), nickel (Ni) and zinc (Zn) bioavailability in a range of soils amended with metal salts and biosolids. Six potential measures of bioavailability were investigated and compared: total metal; 0.04 M EDTA extraction; 0.05 M Ca(NO3)2 extraction; soil solution extracted using rhizon probes; effective solution concentration (CE) determined using diffusive gradients in thin films (DGT); and modelled free ion activities (WHAM 6.0). These were compared to shoot metal concentrations obtained from plants grown in three soils with contrasting properties treated with biosolids and metal salts. The first study involved a wheat seedling (Triticum aestivum) assay carried out under controlled environmental conditions on incubated soils treated with metal salts and biosolids. Results showed that the presence of biosolids resulted in increases of DOC, salinity, Ca and Mg in soil solution as well as total concentrations of Cu and Zn, dry matter was also adversely affected by increased levels of salinity. The addition of biosolids did not significantly alter the extractability or solubility of Cd, Cu, Ni and Zn although concentrations of Cd in shoots were significantly lower in plants grown in biosolids amended soils compared with unamended soils. The second study involved a field experiment that used 20 cm diameter by 30 cm deep soil monoliths of the same three soils treated with metals and biosolids, and perennial ryegrass (Lolium perenne) was grown for 24 months. Results revealed the addition of biosolids significantly increased the amount of DOC, salinity, Ca and Mg in solution. The presence of biosolids also significantly altered the bioavailability of Cd, Cu, Ni and Zn, as measured by soil solution, CE and free ion activity. However, this change had little effect on plant metal uptake. The length of time following treatment application had the greatest effect on soil chemistry and metal availability, resulting in pH decreases and increases in DOC, soil solution salinity, Ca and Mg. The free ion activities of each metal increased with time, as did soil solution Cd and Zn and CE-Cu, with results for Zn indicative of migration through the soil profile with time. Plant uptake of Ni and Zn also changed with time. Nickel concentrations in shoots decreased, while concentrations of Zn in shoots increased. The findings from the two studies demonstrated that biosolids increased the amount of DOC, salinity, Ca and Mg present in soil solution. In the lysimeter study measures of metal availability were affected in soils amended with biosolids, but this did not effect shoot concentrations. The overall predictive strengths of the six potential measures of bioavailability was investigated using results from the previously described experiments and related studies carried out by ESR and Lincoln University using nine different soils amended with combinations of biosolids and metal salts. Of the four
ii metals Ni provided the strongest correlations between metal bioavailability and shoot concentrations, with 0.05 M Ca(NO3)2 extraction giving the strongest relationship for Ni concentrations in shoots (r2 = 0.73). This suggests that the solubility of Ni is highly indicative of shoot concentrations and that Ca(NO3)2 is a robust measure of Ni bioavailability. In 2 addition Ca(NO3)2 provided the best estimate of Zn bioavailability (r = 0.65), and CE-Cd provided the best measure of Cd bioavailability, although it could only describe 47 % of shoot Cd concentration. Results for Cu were typical of previously described studies as assays of Cu availability are almost always poorly correlated with shoot concentrations, with total Cu having the strongest relationship (r2 = 0.34). Methods based on the extractability and solubility of Cu in soils were poor indicators of Cu concentration in shoots. Overall, the addition of biosolids did not alter the outcome of these bioavailability assays, and results indicated that total metal concentrations present in the soils and biosolids matrix, plus length of time since soil treatment, had a greater affect on metal bioavailability.
Keywords: Bioavailability, Cd, Cu, Ni, Zn, biosolids, wheat (Triticum aestivum), ryegrass (Lolium perenne), soil monoliths, pot trials
iii Acknowledgements
I owe this thesis to a number of people, in particular I would like to thank Professor Ron McLaren, Dr Suzie Reichman, Dr Tom Speir and Professor Leo Condron for their alternating supervisory roles - your collective expertise, efforts and critical eye were always appreciated. I would especially like to thank Maui Husdon, Dr Tom Speir, Dr Jacqui Horswell and Dr Val Orchard at ESR Ltd. for their financial support of me and my research through the ESR Doctoral Scholarship Programme; to Nga Pae o te Maramatanga for their financial support through the Mai Doctoral Scholarship Programme, and to the Christchurch City Council for logistical funding provided.
To Lynne Clucas, to whom I owe a big thank you, Roger Creswell, Jason Breitmeyer, Joy Jiao and Neil Smith for their invaluable assistance in everything technical. Field assistance with lysimeter collection, housing and maintenance was provided by Dave Jack, Rob Dijcker, Trevor Hendy, Brendan Malcom, Baiq Dewi Krisnayanti, Debbie Kapal, Robyn Damary- Homan, Dr. Dan Ressler and was greatly appreciated.
To my friends, Dr Gary Houliston (another big thank you) and Tia Neha for statistical assistance and second opinions; to Dr Simon Lambert, Dr Tim Moore, Professor Leo Condron, Dr Janet Bertram, Dr Laurie Steiner, Angela Stuart, Kelly Leers and Peter Cragg for making sure that I wasn’t going too insane or financially destitute.
Finally to my family (Mum, Dad, Regan, Damian and my late grandparents) for not really understanding what it is I do for a living, but continually providing me with their unconditional support.
Amanda
iv Table of Contents
Abstract ...... ii Acknowledgements...... iv Table of Contents...... v List of Tables...... viii List of Figures ...... xi Chapter 1 Literature Review...... 1 1.1 The concept of bioavailability in relation to metals in soils 1 1.1.1 Consequences of excess metals in the soil environment 2 1.2 Soil properties and processes that affect metal bioavailability 4 1.2.1 Total metals in soils 5 1.2.2 pH 5 1.2.3 Organic matter 6 1.2.4 Clays and oxides 9 1.2.5 Redox conditions 9 1.2.6 Rhizosphere processes 10 1.2.6.1 pH changes 10 1.2.6.2 Mycorrhiza and root exudates 11 1.2.6.3 Rhizobia and microbial activity 12 1.3 Metals in plants 13 1.3.1 Mechanisms of metal uptake and transport 13 1.3.2 Toxicity responses to elevated metal concentrations in soils 14 1.3.2.1 Cadmium 15 1.3.2.2 Copper 15 1.3.2.3 Nickel 15 1.3.2.4 Zinc 16 1.4 Methods for determining metal bioavailability in soils 17 1.4.1 Total extractable metals 17 1.4.2 Chelation methods 17 1.4.3 Weak and neutral salt extractants 18 1.4.4 Metals in soil solution (water extractable) 19 1.4.5 Diffuse gradient in thin films 21 1.4.6 Speciation modeling and free ion activity 22 1.5 The use of biosolids as a soil amendment 23 1.5.1 Biosolids research and management in New Zealand 25 Chapter 2 Research Aim and Objectives ...... 27 2.1 Aim 27 2.2 Objectives 27 2.3 Structure 28 Chapter 3 Soil Classification and Methodology ...... 30 3.1 Soil localities and classifications 30 3.2 Characterisation of soils and biosolids 33 3.3 Plant and soil analyses using AAS, GFAAS and ICP-OES 35 3.3.1 Analysis of plant samples 36 3.3.2 Analysis of soil samples for estimates of metal availability 36 3.3.2.1 Total recoverable metals 36 3.3.2.2 EDTA extractions 37
v 3.3.2.3 Calcium nitrate (CaNO3)2 extractions 37 3.3.2.4 Soil solution extractions 37 3.3.2.5 Diffusive gradient in thin-films and DIFS analyses 39 3.3.2.6 Speciation and solid-solution phase modelling using WHAM 6.0 41 Chapter 4 Evaluating the Effects of Biosolids Amendment on the Bioavailability of Cd, Cu, Ni and Zn in Soils Maintained Under Controlled Environmental Conditions... 43 4.1 Introduction 43 4.2 Methods 45 4.2.1 Set-up of incubation trials 45 4.2.2 Seedling bioassay experiments 46 4.2.3 Derivation of critical levels of metals in plant tissues 48 4.2.4 Statistical analyses 48 4.3 Results 48 4.3.1 Changes in soil chemistry 48 4.3.1.1 Major cations, dissolved organic carbon and pH 48 4.3.1.2 Metal extractability 49 4.3.2 Responses of plants 51 4.3.2.1 Estimations of metal bioavailability 54 4.3.2.2 Dry matter production 59 4.3.2.3 Metal concentrations in wheat shoots 63 4.3.3 Effects of biosolids amendments on relationships between soil extraction methods and metal concentrations in shoots 68 4.4 Discussion 70 4.4.1 Influence of biosolids on the availability of metals in soils 70 4.4.2 Responses of plants grown in metal spiked soils in the presence and absence of biosolids 72 4.4.1.1 Dry weight reduction 72 4.4.1.2 Metal concentrations in shoots 73 4.3.2 Estimates of metal bioavailability 75 4.5 Conclusions 77 Chapter 5 Evaluating the Effects of Biosolids and Metal Salt Applications on the Bioavailability of Cd, Cu, Ni and Zn in Soils Managed Under Field Conditions...... 79 5.1 Introduction 79 5.2 Methods 80 5.2.1 Set-up of lysimeter experiment and treatment of soils 80 5.2.2 Soil treatments 81 5.2.3 Harvesting of ryegrass and sampling of soil 83 5.2.4 Analysis of plant and soil samples 83 5.2.5 Derivation of critical levels of metals in plant tissues 84 5.2.6 Statistical analyses 84 5.3 Results 85 5.3.1 Changes in soil chemistry 85 5.3.1.1 Dissolved organic carbon, pH and major anions and cations 85 5.3.1.2 Metal extractability 86 5.3.2 Responses of plants 91 5.3.2.1 Metal concentrations in ryegrass 92 5.3.2.2 Estimations of metal availability 98 5.3.2.3 Effects of biosolids amendments on relationships between soil extraction methods and metal concentration in plants 104 5.4 Discussion 106
vi 5.4.1 Influence of biosolids and time after treatment application on metal availability in soils 106 5.4.2 Responses of plants grown in biosolids amended and unamended metal spiked soils 109 5.4.2.1 Presence of biosolids 109 5.4.2.2 Change in shoot metal concentrations over time 110 5.4.3 Estimates of metal bioavailability 112 5.4.4 Conclusions 115 Chapter 6 Evaluating Methods for Estimating the Bioavailability of Metals Using a Range of New Zealand Soils Amended with Different Rates of Biosolids and Metal Salts...... 117 6.1 Introduction 117 6.2 Methods 120 6.2.1 Set up of ESR and Lincoln University trials 120 6.2.1.1 Field lysimeter study to assess the effects of annual biosolids application to land 120 6.2.1.2 Pasture field trial to assess the effects of biosolids spiked with metal salts on soil properties and plant uptake of metals 121 6.2.1.3 Pasture field trials to assess the effects of metal salt applications on soil chemistry and plant uptake 122 6.2.1.4 Addition of metal spiked biosolids to forest soils 123 6.2.2 Statistical analyses 124 6.3 Results 125 6.3.1 Effects of metal spiking of biosolids and soils on estimates of metal bioavailability 125 6.3.2 Effects of plant species on estimates of bioavailability 129 6.3.3 Estimations of bioavailability using pooled plant species and soil data 136 6.4 Discussion 142 6.4.1 Effects of metal spiking of biosolids and soils on estimates of bioavailability 142 6.4.2 Effects of plant species on estimates of bioavailability 145 6.4.3 Evaluation of six potential measurements of bioavailability using pooled plant species and soil data 148 6.5 Conclusions 150 Chapter 7 General Discussion, Conclusions and Approaches for Further Studies ...... 152 7.1 Summary 152 7.2 Estimates of bioavailability 153 7.3 Influence of biosolids on metal bioavailability 158 7.4 Plant responses 159 7.5 Conclusions 160 7.6 Approaches for further studies 161 References ...... 163 Appendix A . Soil chemistry, metal extractability and speciation modelling results for Chapter 4...... 183 Appendix B . Plant results for Chapter 4 ...... 202 Appendix C . Regression and ANOVA results for DOC, salinity and shoot Cd concentrations for Chapter 4...... 224
vii List of Tables
Table 1.1 Biosolids classification by contaminant levels grades a + b (Cavanagh and Coakley, 2005)...... 26 Table 3.1 Characteristics and locations of experimental soils...... 32 Table 3.2 Calculated MWHC, pH, Al and Fe oxides, CEC and organic carbon values for three New Zealand soils and biosolids sourced from the Bromley wastewater treatment plant used in experiments described in Chapter 4 and 5...... 34 Table 3.3 Calculated values of total base saturation (TBS), exchangeable bases (EB) and percentage of base saturation (BS) for three New Zealand soils and biosolids sourced from the Bromley wastewater treatment plant used in experiments described in Chapters 4 and 5...... 35 Table 3.4 Results of the soil texture analysis determined for three New Zealand soils used in experiments described in Chapters 4 and 5...... 35 Table 4.1 Soil treatments for incubation experiments. Metal were added in the form of hydrated sulphates and biosolids were applied at a rate equivalent to 400 kg N ha-1...... 46 Table 4.2 Significance levels for comparisons of general soil parameters between biosolids amended and unamended metal spiked soils (n = 180). Differences were considered significant at P<0.05 and highlighted in bold...... 49 Table 4.3 Significance levels for comparisons of soil parameters between all treated soils sampled after two weeks and 24 weeks of incubation, n = 90. Differences were considered significant at P<0.05 and highlighted in bold...... 49 Table 4.4 Significance values for comparisons of six potential measures of metal bioavailability between biosolids amended and unamended metal spiked soils (log transformed data, n = 180). Differences were considered significant at P<0.05 and highlighted in bold...... 50 Table 4.5 Significance levels for comparisons of six potential assays of metal bioavailability between treated soils incubated for two weeks and soils incubated for 24 weeks (log transformed data, n = 90). *Total metal was only measured on one occasion (24 weeks incubated soils). Differences were considered significant at P<0.05 and highlighted in bold...... 50 Table 4.6 (A) Significance levels for comparisons between shoot metal concentrations in plants grown in biosolids-amended and unamended metal spiked soils (log transformed data, n = 35). (B) Significance levels for comparisons between two and 24 weeks incubation (log transformed data, n = 35). Differences were considered significant at P<0.05 and highlighted in bold...... 67 Table 4.7 Significance levels for comparisons between shoot metal concentrations in plants grown in biosolids-amended and unamended metal treated soils, in order of increasing metal treatment levels (log transformed data, n = 70). Differences were considered significant at P<0.05 and highlighted in bold...... 67 Table 5.1 Soil treatments for field lysimeter experiment. Metals were added in the form of sulphates and biosolids were applied at a rate equivalent to 400 kg N ha-1 ...... 82 Table 5.2 Significance levels for comparisons of general soil parameters between pooled results for biosolids-amended (BLM) and unamended (LM) metal spiked soils (n = 27). Differences were considered significant at P<0.05 and highlighted in bold (*data log10-transformed)...... 85 Table 5.3 Mean soil chemistry characteristics from pooled treated soils, sampled six months, 12 months and 24 months after treatment application (values are means, n = 63)...... 86
viii Table 5.4 Effect of time calculated from residual deviance (variation due to treatment effects removed) on soil chemistry charactersitics. Effects were considered significant at P<0.05 and highlighted in bold (*data log10-transformed). P- value for the T distribution is also included...... 86 Table 5.5 Significance levels for comparisons of six potential measures of metal bioavailability between pooled sampling results for biosolids amended (BLM) and unamended (LM) metal spiked soils (log transformed data, n = 27). Differences were considered significant at P<0.05 and highlighted in bold...... 87 Table 5.6 Means for six potential measures of Cd bioavailability taken from pooled soil results from each sampling occasion (n = 63 for each sampling occasion)...... 88 Table 5.7 Effect of time as calculated from residual deviance (variation due to treatment effects removed) on six measures of Cd bioavailability. Effects were considered significant at P<0.05 and highlighted in bold (*data log10- transformed). P-value for the T distribution is also included...... 88 Table 5.8 Means for six potential measures of Cu bioavailability taken from pooled soil results from each sampling occasion (n = 63 for each sampling occasion)...... 89 Table 5.9 Effect of time as calculated from residual deviance (variation due to treatment effects removed) on six measures of Cu bioavailability. Effects were considered significant at P<0.05 and highlighted in bold (*data log10- transformed). P-value for the T distribution is also included...... 89 Table 5.10 Means for six potential measures of Ni bioavailability taken from pooled soil results for each sampling occasion (n = 63 for each sampling occassion)...... 89 Table 5.11 Effect of time as calculated from residual deviance (variation due to treatment effects removed) on six measures of Ni bioavailability. Effects were considered significant at P<0.05 and highlighted in bold (*data log10-transformed). P- value for the T distribution is also included...... 90 Table 5.12 Means for six potential measures of Zn bioavailability taken from pooled soil results from each sampling occasion (n = 63 for each sampling occasion)...... 90 Table 5.13 Effect of time calculated from residual deviance (variation due to treatment effects removed) on six measures of Zn bioavailability. Effects were considered significant at P<0.05 and highlighted in bold (*data log10- transformed). P-value for the T distribution is also included...... 90 Table 5.14 Significance levels for comparisons between shoot metal concentrations in plants grown in biosolids-amended (BLM) and unamended (LM) metal spiked soils harvested after treatment application (log transformed data, n = 45, 54, 56 and 59 for 6, 12, 18 and 24 months, respectively). Differences were considered significant at P<0.05 and highlighted in bold...... 97 Table 5.15 Effect of time calculated from residual deviance (variation due to treatment effects removed) on plant uptake of metals. Effects were considered significant at P<0.05 and highlighted in bold (*data log10-transformed). P-value for the T distribution is also included...... 98 Table 5.16 Significance levels obtained from grouped regression analyses for relationships determined between Cd concentrations in ryegrass and five potential measures of Cd availability in biosolids-amended (BLM) and unamended (LM) metal spiked soils. Differences were considered significant at P<0.05 and highlighted in bold (n = 27)...... 105 Table 5.17 Significance levels obtained from grouped regression analyses for relationships determined between Cu concentrations in ryegrass and five potential measures of Cu availability in biosolids-amended (BLM) and unamended (LM) metal spiked soils. Differences were considered significant at P<0.05 and highlighted in bold (n = 27)...... 105
ix Table 5.18 Significance levels obtained from grouped regression analyses for relationships determined between Ni concentrations in ryegrass and five potential measures of Ni availability in biosolids-amended (BLM) and unamended (LM) metal spiked soils. Differences were considered significant at P<0.05 and highlighted in bold (n = 27)...... 106 Table 5.19 Significance levels obtained from grouped regression analyses for relationships determined between Zn concentrations in ryegrass and five potential measures of Zn availability in biosolids-amended (BLM) and unamended (LM) metal spiked soils. Differences were considered significant at P<0.05 and highlighted in bold (n = 27)...... 106 Table 6.1 Coefficient of determination (r2) for each soil amendment group from relationships determined between plant concentration and six potential measures of Cd bioavailability. The highest scoring significant r2 value (P<0.05) for each group is highlighted in red. All data was log10 transformed for analysis. * P-values were greater than 0.05...... 127 Table 6.2 Coefficient of determination (r2) for each soil amendment group from relationships determined between plant concentration and six potential measures of Cu bioavailability. The highest scoring significant r2 value (P<0.05) for each group is highlighted in red. All data was log10 transformed for analysis. * P-values were greater than 0.05...... 127 Table 6.3 Coefficient of determination (r2) for each soil amendment group from relationships determined between plant concentration and six potential measures of Ni bioavailability. The highest scoring significant r2 value (P<0.05) for each group is highlighted in red. All data was log10 transformed for analysis. * P-values were greater than 0.05...... 128 Table 6.4 Coefficient of determination (r2) for each soil amendment group from relationships determined between plant concentration and six potential measures of Zn bioavailability. The highest scoring significant r2 value (P<0.05) for each group is highlighted in red. All data was log10 transformed for analysis. * P-values were greater than 0.05...... 129 Table 6.5 Significance levels and differences between plant species as determined from comparisons of relationships (gradients) between metal concentration in shoots of 20 day old wheat seedlings and ryegrass, and six potential measures of metal bioavailability. All data was log10 transformed for analysis. Differences were considered significant at P<0.05 and highlighted in bold...... 135 Table 7.1 Range of significant (P<0.05) coefficients of determination for six potential measures of bioavailability assessed in three related studies. Total varation for each method and metal is derived from the highest r2 value minus the lowest r2 value...... 155
x List of Figures
Figure 3.1 Localities of soils used in experiments described in Chapters 4, 5 and 6...... 31 Figure 3.2 Soil solution extraction using Rhizon MOM type probes from soils used in the wheat seedling bioassays described in Chapters 4 and 6...... 38 Figure 3.3 Washing of DGT device with de-ionised water after 24 hour placement on soil wetted to 90 ± 10% MWHC and allowed to equilibrate. Photo B shows removal of filter and diffusive gel, and photo C shows placement of resin gel in 10 ml of 0.01 M HNO3...... 41 Figure 4.1 A. Appearance of wheat seedlings after 20 days of growth in untreated (control) soil previously incubated for six months at 50% of maximum water holding capacity. Scale is 10 centimetres...... 51 Figure 4.2 A. Appearance of wheat seedlings after 20 days of growth in all three soils treated with a combination of biosolids and 5, 200, 150 and 300 mg kg-1 of Cd, Cu, Ni and Zn (biosolids and medium metals), and incubated for 24 weeks at field capacity. B. Close-up of seedlings grown in HK soil showing first signs of chlorosis in some of the leaves as a result of soil treatment. Scale is 10 centimetres...... 52 Figure 4.3 A. Appearance of wheat seedlings after 20 days of growth in all three soils treated with a combination of biosolids and 10, 750, 300 and 1000 mg kg-1 of Cd, Cu, Ni and Zn, and incubated for 24 weeks at field capacity. B. Appearance of seedlings in all three soils grown in 10, 750, 300 and 1000 mg kg-1 of Cd, Cu, Ni and Zn without the addition of biosolids. Scale is 10 centimetres...... 53 Figure 4.4 Relationships determined between shoot Cd concentrations and Cd concentrations in soil obtained using various methods: (A) Total Cd, (B) EDTA-extractable Cd, (C) Ca(NO3)2 extractable Cd, (D) soil solution, (E) effective concentration, (F) free ion activity. A linear regression line is fitted for each plot. n = 175...... 55 Figure 4.5 Relationships determined between shoot Cu concentrations and Cu concentrations in soil obtained using various methods: (A) Total Cu, (B) EDTA-extractable Cu, (C) Ca(NO3)2 extractable Cu, (D) soil solution, (E) effective concentration, (F) free ion activity. A linear regression line is fitted for each plot. n = 175...... 56 Figure 4.6 Relationships determined between shoot Ni concentrations and Ni concentrations in soils obtained using various methods: (A) Total Ni, (B) EDTA-extractable Ni, (C) Ca(NO3)2 extractable Ni, (D) soil solution , (E) effective concentration, (F) free ion activity. A linear regression line is fitted for each plot. n = 175...... 57 Figure 4.7 Relationships determined between shoot Zn concentrations and concentrations of Zn in the soil obtained using various methods: (A) Total Zn, (B) EDTA- extractable Zn, (C) Ca(NO3)2 extractable Zn, (D) soil solution, (E) effective concentration, (F) free ion activity. A linear regression line is fitted for each plot. n = 175...... 58 Figure 4.8 Mean dry weights of wheat seedlings grown in biosolids amended and unamended metal treated soils incubated for two weeks and 24 weeks at field capacity. Bars denote the SEM, n = 89 for both trials...... 60 Figure 4.9 A-D. Relationships determined between plant dry weight and corresponding Ca(NO3)2 extractable metal concentrations in biosolids amended and unamended metal treated soils, (A) = Cd, (B) = Cu, (C) = Ni, (D) = Zn, n = 179...... 61
xi Figure 4.10 Relationship determined between plant dry weight and the salinity of the soil solution, n = 179...... 62 Figure 4.11 Mean Cd concentrations in wheat shoot tissue (mg kg-1 dry weight) grown in biosolids amended and unamended metal spiked soils incubated for two weeks and 24 weeks. The critical level (CL) for Cd as depicted by the red line is 43 mg kg-1 (Macnicol and Beckett, 1985). Bars denote the SEM, n = 89 for both trials...... 64 Figure 4.12 Mean Cu concentrations in wheat shoots (mg kg-1 dry weight) grown in biosolids amended and unamended metal spiked soils incubated for 2 weeks and 24 weeks. The critical levels (CL) for Cu as depicted by the red line, is 17 mg kg-1 (Macnicol and Beckett, 1985). Bars denote the SEM, n = 89 for both trials...... 65 Figure 4.13 Mean Ni concentrations in wheat shoots (mg kg-1 dry weight) grown in biosolids amended and unamended metal spiked soils incubated for 2 weeks and 24 weeks. The critical levels (CL) for Ni as depicted by the red line, is 46 mg kg-1 (Macnicol and Beckett, 1985). Bars denote the SEM, n = 89 for both trials...... 66 Figure 4.14 Mean Zn concentrations in wheat shoot tissue (mg kg-1 dry weight) grown in biosolids amended and unamended metal spiked soils incubated for 2 weeks and 24 weeks. The critical levels (CL) for Zn as depicted by the red line, is 224 mg kg-1 (Macnicol and Beckett, 1985). Bars denote the SEM, n = 89 for both trials...... 66 Figure 4.15 Metal bioavailability estimates as determined by Ca(NO3)2 extractions as affected by biosolids amendment of three soils spiked with metal salts: (A) results for Cd; (B) results for Cu; (C) results for Ni; (D) results for Zn .n = 71 (Biosolids + MS) and n = 72 (MS)...... 69 Figure 4.16 Relationship between shoot Zn concentration and effective concentration as affected by amendment with biosolids, n = 71 (Biosolids + MS) and n = 72 (MS)...... 70 Figure 5.1 Map view and photo of field lysimeter set-up. Locations of soil and treatments within the trench are also shown...... 81 Figure 5.2 Apperance of ryegrass in two of the soil lysmeters six months after treatment application (MM = medium metal treatment)...... 92 Figure 5.3 Mean Cd concentrations in ryegrass shoot tissue (mg kg-1 dry weight) grown in biosolids amended and unamended metal spiked soils managed under field conditions over 24 months. The critical level (CL) for Cd as depicted by the red line is 30-35 mg kg-1 (Macnicol and Beckett, 1985). Bars denote the SEM, n = 45, 54, 56 and 59 for 6, 12, 18 and 24 months, respectively...... 94 Figure 5.4 Mean Cu concentrations in ryegrass shoot tissue (mg kg-1 dry weight) grown in biosolids amended and unamended metal spiked soils managed under field conditions over 24 months. The critical level (CL) for Cu as depicted by the red line is 30-35 mg kg-1 (Macnicol and Beckett, 1985). Bars denote the SEM, n = 45, 54, 56 and 59 for 6, 12, 18 and 24 months, respectively...... 95 Figure 5.5 Mean Ni concentrations in ryegrass shoot tissue (mg kg-1 dry weight) grown in biosolids amended and unamended metal spiked soils managed under field conditions over 24 months. The critical level (CL) for Ni as depicted by the red line is 130-220 mg kg-1 (Macnicol and Beckett, 1985). Bars denote the SEM, n = 45, 54, 56 and 59 for 6, 12, 18 and 24 months, respectively...... 96 Figure 5.6 Mean Zn concentrations in ryegrass shoot tissue (mg kg-1 dry weight) grown in biosolids amended and unamended metal spiked soils managed under field conditions over 24 months. The critical level (CL) for Zn as depicted by the red
xii line is 370-560 mg kg-1 (Macnicol and Beckett, 1985). Bars denote the SEM, n = 45, 54, 56 and 59 for 6, 12, 18 and 24 months, respectively...... 97 Figure 5.7 Relationships determined between soil extractable Cd and shoot Cd concentration using various extractants; (A) Total Cd, (B) EDTA-extractable Cd (C) Ca(NO3)2 extractable Cd, (D) soil solution, (E) effective concentration, (F) free ion activity. A linear regression line is fitted for each plot. n = 157...... 100 Figure 5.8 Relationships determined between soil extractable Cu and shoot Cu concentration using various methods; (A) Total Cu, (B) EDTA-extractable Cu (C) Ca(NO3)2 extractable Cu, (D) soil solution, (E) effective concentration, (F) free ion activity. A linear regression line is fitted for each plot. n = 157...... 101 Figure 5.9 Relationships determined between soil extractable Ni and shoot Ni concentration using various extractants; (A) Total Ni, (B) EDTA-extractable Ni (C) Ca(NO3)2 extractable Ni, (D) soil solution, (E) effective concentration, (F) free ion activity. A linear regression line is fitted for each plot. n = 157...... 102 Figure 5.10 Relationships determined between soil extractable Zn and shoot Zn concentration using various extractants; (A) Total Zn, (B) EDTA-extractable Zn (C) Ca(NO3)2 extractable Zn, (D) soil solution, (E) effective concentration, (F) free ion activity. A linear regression line is fitted for each plot. n = 157...... 103 Figure 6.1. Lysimeter set-up at Lincoln University to assess the effects on leachate quality, soil chemistry and herbage uptake of metals from annual application of biosolids...... 121 Figure 6.2 Photograph of biosolids spiked with metal salts field trial, Top photo: Lincoln University trial set up in 1997, Bottom photo: metal salt trial set-up in 2006, Nelson, South Island, New Zealand...... 123 Figure 6.3 Relationships determined between 20 day old wheat seedlings, pasture ryegrass concentrations and Cd concentrations obtained using various methods; (A) Total Cd, (B) EDTA-extractable Cd, (C) Ca(NO3)2 extractable Cd, (D) soil solution, (E) effective concentration, (F) free ion activity (n = 272 for wheat and n = 188 for ryegrass)...... 131 Figure 6.4 Relationships determined between 20 day old wheat seedlings, pasture ryegrass concentrations and Cu concentrations obtained using various methods; (A) Total Cu, (B) EDTA-extractable Cu, (C) Ca(NO3)2 extractable Cu, (D) soil solution, (E) effective concentration, (F) free ion activity for 20 day old wheat seedlings and ryegrass( n = 439 for wheat and n = 277 for ryegrass)...... 132 Figure 6.5 Relationships determined between 20 day old wheat seedlings, pasture ryegrass concentrations and Ni concentrations obtained using various methods; (A) Total Ni, (B) EDTA-extractable Ni (C) Ca(NO3)2 extractable Ni, (D) soil solution, (E) effective concentration, (F) free ion activity (n = 295 for wheat and n = 208 for ryegrass)...... 133 Figure 6.6 Relationships determined between 20 day old wheat seedlings, pasture ryegrass concentrations and Zn concentrations obtained using various methods; (A) Total Zn, (B) EDTA-extractable Zn (C) Ca(NO3)2 extractable Zn, (D) soil solution, (E) effective concentration, (F) free ion activity (n = 439 for wheat and n = 220 for ryegrass)...... 134 Figure 6.7 Relationships determined between shoot Cd concentration and soil Cd concentrations obtained using various methods; (A) Total Cd, (B) EDTA- extractable Cd, (C) Ca(NO3)2 extractable Cd, (D) soil solution, (E) effective concentration, (F) free ion activity. A linear regression line is fitted for each plot. n = 399...... 138 Figure 6.8 Relationships determined between shoot Cu concentration and soil Cu concentrations obtained using various methods; (A) Total Cu, (B) EDTA- extractable Cu, (C) Ca(NO3)2 extractable Cu, (D) soil solution, (E) effective
xiii concentration, (F) free ion activity. A linear regression line is fitted for each plot. n = 592...... 139 Figure 6.9 Relationships determined between shoot Ni concentration and soil Ni concentrations obtained using various methods; (A) Total Ni, (B) EDTA- extractable Ni, (C) Ca(NO3)2 extractable Ni, (D) soil solution, (E) effective concentration, (F) free ion activity. A linear regression line is fitted for each plot. n =425...... 140 Figure 6.10 Relationships determined between shoot Zn concentration and soil Zn concentrations obtained using various methods; (A) Total Zn, (B) EDTA- extractable Zn, (C) Ca(NO3)2 extractable Zn, (D) soil solution, (E) effective concentration, (F) free ion activity. A linear regression line is fitted for each plot. n = 594...... 141
xiv Chapter 1 Literature Review
1.1 The concept of bioavailability in relation to metals in soils
The definition of bioavailability of elements and compounds varies according to the subject of the study. The concept of bioavailability is driven by the organisms or solution phase used for the measurement of the contaminant availability. For example, the bioavailable fraction is often defined as the fraction of contaminant extractable in a chemical reagent based on correlation with the total contaminant uptake by plants (Allen et al., 2001; Alloway et al., 1988; Chojnacka et al., 2005; Fairbrother et al., 2007; McBride et al., 1997a; McLaughlin et al., 2000). In these cases bioavailability is largely a measure of contaminant solubility (Chaney et al., 2000; Degryse et al., 2009a; Hooda, 2007).
Considerable debate exists in the literature on the precise definition of what constitutes a bioavailable metal pool (Hooda, 2007). Metal bioavailability has been linked to metal ion activity in soil solution, the exchangeable metal fraction and more recently to the concentration of metals that cause ecotoxicity (Antunes and Kreager, 2009; Antunes et al., 2007; Hooda, 2007; Lofts et al., 2007; Menzies et al., 2007; Rooney et al., 2006; Rooney et al., 2007b; Sauve, 2001; Sauve, 2003; Smolders et al., 2009; Speir et al., 2007). However, there is still no general consensus among the scientific community on the components involved in determining bioavailability. This may be due to the many inconsistencies between the bioavailable fraction measured and the uptake of metals by plants and the largely undescribed processes that may influence bioavailability occurring in the rhizosphere (Basar, 2009; Gregory, 2006; Jones et al., 2004; Menzies et al., 2007).
Even though it is widely acknowledged that metal bioavailability, rather than total concentration, is critical when accurately assessing the risk associated with soil contamination, debate continues on what constitutes a bioavailable pool and how to best predict bioavailability, especially in relation to crop plants (Alloway, 1995b; Alloway et
1 al., 1988; Gray and McLaren, 2006; McLaughlin, 2002; McLaughlin et al., 2000; Nolan et al., 2005; Oliver et al., 2004; Speir et al., 2003b).
1.1.1 Consequences of excess metals in the soil environment
Unlike organic contaminants which degrade over time (e.g. pesticides and hydrocarbon contaminants), metals can remain in the environment and amounts can only be reduced by transportation elsewhere such as uptake and removal by plants (phytoremediation), which has been observed in many New Zealand and international cases (Adams et al., 2001; Almas et al., 2004; Almas et al., 2007; Chaignon et al., 2003; Chapman, 2008; Georgieva et al., 2002; Gray et al., 1999c; Gray et al., 2002; Knights et al., 2001; McBride, 1995; McLaren et al., 2003; Nwachukwu and Pulford, 2008). The accumulation of metals in soil, particularly Cd, Cu, Ni and Zn, is of concern in agricultural production systems due to the potential threat of adversely affecting food quality (safety and marketability), crop growth (through phytotoxicity), or environmental health (soil flora and fauna). Agricultural exports are increasingly marketed internationally on the basis of an environmentally sound image so the issue of managing and regulating metal contamination of agricultural soils has become an important issue (McLaren et al., 2003; 2004; McLaughlin et al., 2000; McLaughlin et al., 2007; Speir et al., 2003b; Warne et al., 2008a; Warne et al., 2008b; Zarcinas et al., 2004).
Cadmium occurs naturally in soils, however for plants and animals Cd is a non-essential element that is bioaccumulating and is toxic at relatively low concentrations (Webb, 1975; Webb and Daniel, 1975). Biological effects resulting from excess Cd in soils has been extensively examined, particularly in relation to phosphate fertilizer applications (Bolan and Duraisamy, 2003; Gray et al., 1999c; Loganathan et al., 2003; Longhurst et al., 2004; McLaughlin, 2002; McLaughlin et al., 1996; McLaughlin et al., 1995; Nicholson et al., 2006; Roberts and Longhurst, 2002; Roberts et al., 1994). Increased concentrations of Cd in agricultural land from amendments such as mineral fertilizers and biosolids have led to enhanced Cd uptake by pasture plants. The consequence of this is potentially significant for Cd absorption by grazing livestock (Roberts et al., 1994). In
2 soils, Cd is a highly mobile metal in relation to leaching and availability to plants (McLaughlin, 2002; McLaughlin et al., 2000). Cadmium contamination of soil is also typically associated with Cu and especially Zn contamination, as most of the common Cd containing agricultural and industrial products added to soils also contains these elements (McLaughlin et al., 2000).
Copper, Ni and Zn, unlike Cd, are essential elements required in small amounts for the healthy functioning of animal and plant metabolic systems (Alloway et al., 1990; Bolan et al., 2003; Palmer and Guerinot, 2009; Reichman, 2002). However, excessive Cu is toxic to soil microorganisms, plants and grazing animals (Bolan et al., 2003; Giller et al., 2009; Speir et al., 2007; Vogeler et al., 2008) and in plants Cu toxicity causes foliar interveinal chlorosis, with the leaves becoming necrotic with increasing Cu exposure (Reichman et al., 2006). Moreover, elevated Cu soil concentrations have been shown to decrease the microbial activity in soils (Antunes et al., 2007; Sauve et al., 1997; Vogeler et al., 2008). This reduction of microbial activity as a result of Cu toxicity also negatively impacts upon the health of the plant as soil microbes and functioning plant systems are intricately linked (Cloutier-Hurteau et al., 2008; Dumestre et al., 1999; Giller et al., 2009; Khan et al., 2005).
The toxicity of Ni is almost always less than Cu, Zn and especially Cd (McLaughlin et al., 2000). Documented cases of Ni toxicity resulting from excess Ni present in the soil have resulted predominantly from practices such as smelting, ore refining, mining, combustion of fossil fuels and long term application of biosolids to agricultural land (Antoniadis and Alloway, 2002a; Antoniadis and Alloway, 2003b; Doig and Liber, 2007; Marschner and Marschner, 1995; McNear et al., 2007; Tye et al., 2004). Nickel plays an important role in lipid metabolism, hematopoiesis and several other biological functions in some organisms (Kozlova et al., 2009). Toxic responses to excess Ni in soils are less clear than those described for Cd, Cu or Zn (Kozlova et al., 2009; Marschner and Marschner, 1995). However, the most frequently documented effect of Ni toxicity is foliar chlorosis and in some instances decreased root growth (Himmelbauer et al., 2005; Marschner and Marschner, 1995; Molas and Baran, 2004). Research into the occurrence of Ni in soils and relationships between plant and soil microorganisms has focused
3 primarily on plants, especially hyperaccumulating species (i.e. Thlaspi species, Penny- crest) grown on serpentine soils and their role in promoting plant growth and Ni accumulation (Himmelbauer et al., 2005; Idris et al., 2006; Ma et al., 2009; Puschenreiter et al., 2005; Wenzel et al., 2003).
Contamination of soil by excess Zn occurs primary as a result of industrial activities such as mining and smelting (Almas et al., 2006; Almas et al., 2007; Chen et al., 2009; Degryse and Smolders, 2006; Gutierrez-Ruiz et al., 2007) or the long term application of biosolids to agricultural land (Antoniadis, 2008; Collins et al., 2003a; Hirsch et al., 1993; Horswell et al., 2006; Kim et al., 2007). However, occurrences of Zn toxicities in plants are far less wide spread than reported cases of Zn deficiency (Alloway, 2009; Broadley et al., 2007). Zinc is the most common crop micro deficiency, especially in high pH soils containing low concentrations of Zn (Alloway, 2009). In cases where Zn is present in soils in biologically toxic concentrations, symptoms of Zn toxicity typically include a general chlorosis of younger leaves (Reichman, 2002), as well as an overall decrease in leaf size compared with unaffected plants (Broadley et al., 2007; Marschner and Marschner, 1995; Reichman, 2002).
1.2 Soil properties and processes that affect metal bioavailability
Bioavailability and mobility of metals in soils are controlled by a series of soil properties and processes such as, metal speciation, soil pH, soil texture, organic matter content, plant type and agronomic management (Chojnacka et al., 2005; Lair et al., 2007; Lair et al., 2006; McBride et al., 1997a; McLaughlin et al., 2000). Metals present in soils can occur in several fractions such as; soil solution, exchangeable, sorbed and organically bound, bound and occluded onto oxides and clay minerals, residual, and within the primary lattice phase of minerals (Alloway, 1990; McBride, 1994). Plants and soil organisms are unable to access the total pool of metals present in the soil, of these metal fractions, the most immediately bioavailable pool of metals are metals present in soil
4 solution, with other metal fractions being less available (Marschner and Marschner, 1995; Reichman, 2002).
1.2.1 Total metals in soils
Metals and metalloids occur in varying concentrations in all soils, most are essential at some levels for biological life such as Co, Cu, Mn, Ni, Se and Zn (low levels). Toxic metals, is a term that usually refers to those metals with non-essential biological roles, such as As, Cd, Pb, Hg, Tl and U (Alloway, 1995b; Alloway et al., 1990). The total concentration of metal in a soil includes all fractions of metals from those present in soil solution, to exchangeable amounts present on organic, clay and oxide surfaces, and to those occluded in the mineral matrices, from which rates of desorption are relatively slow (McBride, 1994). While total metal concentration in soils is still frequently used to estimate bioavailability, this assessment is considered insensitive to adequately reflect relevant metal availability (McBride et al., 1997b; McLaughlin et al., 2000; Sauve, 2003)
1.2.2 pH
Soil pH is regarded as the fundamental independent variable that controls ion exchange, dissolution/precipitation, redox reactions, adsorption, and complexation reactions occurring within the soil (McBride, 1994). Soil solution pH is intimately involved in the complex equilibria between metal speciation, solubility, adsorption on colloids and sites available for cation exchange (McBride, 1994; Reichman, 2002; Sauve, 2001). The binding of metal cations is strongly affected by pH, firstly due to the competition between protons and metal cations for humic binding sites, and secondly because of cation hydrolysis (Tipping, 1998; Tipping et al., 2003a).
Soil pH is recognised as the major factor controlling Cd and especially Zn availability, (Antonladis et al., 2008; Carrillo-Gonzalez et al., 2006; Gray and McLaren, 2006; McBride et al., 1997a; McBride et al., 1997b; Rooney et al., 2007b). Studies examining the factors that influence Cd availability revealed that pH was the dominant soil variable
5 affecting solution Cd (Andrews et al., 1996; Gray and McLaren, 2006; Gray et al., 2003; Gray et al., 1999b; 1999d; Loganathan et al., 2003; Rachou et al., 2004). In addition, organic matter, CEC and total soil Cd were also found to be important factors determining Cd bioavailability (Gray and McLaren, 2006).
In many studies, Ni solubility and toxicity in soils was positively related to soil pH and total Ni content, while negatively related to total carbon content (Antoniadis and Alloway, 2002a; Antoniadis and Alloway, 2002b; McLaughlin et al., 2000; Rooney et al., 2007b). While pH has clearly been established as the key soil property controlling the solubility of Cd and Zn, the circumstances for Ni differs as organic matter has a more dominant role in influencing Ni solubility than for Cd or Zn (Sauve and Parker, 2005) and plant uptake of Ni (Hough et al., 2003; Hough et al., 2005).
The effect of pH on Cu speciation is less well defined and while some studies examining the influence of pH on the availability and toxicity of these metals found that soil solution pH does affect Cu solubility and speciation (Chaignon et al., 2003), other studies have found little direct relationship between pH and Cu availability (Sauve and Parker, 2005; Sauve et al., 1997). The reason is that Cu has a strong affinity for organic matter (Burton et al., 2005a; Burton et al., 2005b; McBride, 1994) and organic matter, especially dissolved organic matter is a more important determinant of Cu solubility and bioavailability than pH (Amery et al., 2007; Amery et al., 2008; Reichman, 2002; Sarathy and Allen, 2005).
1.2.3 Organic matter
Organic matter is a fundamental component of the soil and influences biological, chemical and physical soil properties (McBride, 1994; Stevenson, 1994). In particular, organic matter is involved the retention and release of nutrients (i.e. Ca, Cu, Fe, K, Mg, Mn) by cation exchange and adsorption of potentially toxic organic compounds, such as pesticides and industrial wastes (McBride, 1994).
6 Organic matter in soils is commonly and collectively referred to as humic substances, and this can be further divided into humic acid, fulvic acid, and humin based on solubility (McBride, 1994). Humic and fulvic acids contribute to the majority of cation binding properties in soils (Tipping, 2005) and dissolved organic carbon (DOC) is the main source of fulvic acids which are known to increase the carrying capacity of soil solutions for strongly organic complexing metals such as Cu (Amery et al., 2007; Ashworth and Alloway, 2007; Bolan and Duraisamy, 2003; Cattani et al., 2006; Chaignon et al., 2003; Sauve et al., 1997), and in some instances Ni (Antoniadis and Alloway, 2002b; Ashworth and Alloway, 2004; Doig and Liber, 2007). Metal ions can be complexed by the COO- and COOH- groups present in both DOC and solid organic matter to form stable complexes (McLaren and Crawford, 1973a; Stevenson, 1994). Consequently, the amount of organic matter, in particular DOC increases the opportunity for forming stable organo- metal complexes (McBride, 1994). In general these stable soluble organo-metal complexes are considered largely unavailable for plant uptake (Reichman, 2002), and whilst the amount of metal in solution may increase, metals, especially Cu may be less available as a result of complexing with DOC (Amery et al., 2008; Kalis et al., 2006).
Several studies have shown that the presence of DOC does effect the solubility and bioavailability of Cd (Antoniadis and Alloway, 2002b; Antonladis and Tsadilas, 2007; Gray and McLaren, 2006; Gray et al., 1999a; McLaughlin, 2002; McLaughlin et al., 2006). A study by Antoniadis and Alloway (2002b) found that increased amounts of DOC from biosolids application to land were correlated with an increase in Cd solubility as well as plant uptake of Cd. In contrast, the case of Cd applied to soils as Cd contained in biosolids, Cd was shown to be less available for plant uptake than the more water soluble forms, such as Cd salts (McLaughlin et al., 2006). Additionally, a study by Antonladis and Tsadilas (2007), observed that the complexation of Cd by DOC and retention in organic matter were less likely to occur in the presence of Cu, Ni or Zn as Cd does not compete for these sites as well as these metals. Thus while DOC plays an important role in determining the solubility and availability of Cd, it appears to be more influenced by the presence of other cations such as Ca (Antonladis and Tsadilas, 2007; Antonladis et al., 2008; Gray and McLaren, 2006)
7 Copper forms strong coordination complexes with organic matter, hence Cu present in soils is relatively immobile, with a large proportion associated with organic matter in the solid phase (Ashworth and Alloway, 2007; McBride et al., 1997b; McLaren and Crawford, 1973a; Reichman, 2002; Stevenson, 1994). In experiments carried out by Burton et al. (2005a and 2005b) it was revealed that Cu was sorbed predominantly to soil organic matter and Cu interactions with DOC were shown to influence the sorption- desorption process that primarily control Cu behaviour in soil (McBride, 1995; McLaren and Crawford, 1973b). Sauve et al. (1997) also confirmed the importance of organic matter in determining Cu solubility and bioavailability with a study on Cu contamination of a range of soils, showing that greater than 90% of Cu was bound organically irrespective of the soil pH.
Nickel ions also form strong coordination complexes with organic matter, because of the metal’s relatively high electronegativity which is second only to Cu in soil environments (McBride, 1994). Information regarding Ni availability in soils in relation to plant uptake or toxicity is substantially less than for Cd, Cu or Zn (McIlveen and Negusanti, 1994; Weng et al., 2003). Studies that have documented the effects organic matter, specifically DOC have had on Ni solubility and bioavailability revealed that complexation by DOC did increase Ni availability to plants and mobility through soil (Antoniadis and Alloway, 2002a; 2002b; Ashworth and Alloway, 2004). However, pH and competitive absorption are thought to be equally influencing of Ni solubility and bioavailability (Antonladis and Tsadilas, 2007; Ponizovsky et al., 2008; Puschenreiter et al., 2005).
Zinc solubility is also affected by organic matter and readily forms complexes with DOC, although it does not compete as well as Ca, Cu or Ni for these sites (McBride et al., 1997a). Zinc uptake in plants as well as Zn mobility down through the soil profile has also been correlated with the amount of DOC in soil solution (Antoniadis and Alloway, 2002a; Antoniadis and Alloway, 2002b; Ashworth and Alloway, 2004). However in these studies and several others the concentration and species of Zn in solution was found to be more strongly related to pH than on the amount of DOC (Antonladis et al., 2008; Broadley et al., 2007; Imtiaz et al., 2006; Lock and Janssen, 2001).
8 While organic matter, especially DOC clearly affects the behaviour of metals in soils, the amount of DOC in soil solution is most likely controlled by the amount of oxides and clays present in the soil. The presence of these other soil constituents, are known to form highly stable complexes with organic matter (McBride, 1994) and the influence of these on metal behaviour in the soil are described in the following sections.
1.2.4 Clays and oxides
Cadmium, Ni, Cu and Zn in soils are usually present in trace amounts, hence sorption processes (often referred to as specific absorption or chemi-adsorption) are generally accepted as the primary mechanisms responsible for controlling the solubility of these metals (McBride et al., 1997a; McBride, 1994). Specific adsorption usually takes place on the surfaces of clays and oxides (principally of Al, Fe, Mn and Si), with the clay fraction (commonly defined as the portion of fraction size < 2µm) usually the most influential for cation adsorption reactions, due to its high surface area, hydroxyl groups and large number of available binding sites (McBride, 1994).
Trace metals, such as Cd and Ni can also be co-precipitated or incorporated into the oxyhydroxide minerals, and in extreme cases (e.g. severe industrial pollution, or geothermal environments), precipitation of metals as distinct minerals can occur (Alloway, 1995a; McBride, 1994).
1.2.5 Redox conditions
The reduction/oxidation conditions in a soil can strongly affect the pH and speciation, and hence toxicity of metals in solution, and this is particularly well documented for As, Sb, Cr, Fe and Mn, which are several times more biologically toxic and mobile in their most reduced form (Allen et al., 2001; Alloway, 1995b; Horswell and Speir, 2006; Oorts et al., 2008). However Cd, Cu, Ni and Zn in soils are mostly present in the divalent form, as the monovalent forms are highly unstable and hence these metals are not significantly
9 reduced under low redox conditions (Reichman, 2002; Whitehead, 2000). Thus redox conditions are considered to play a much smaller role in the solubility and bioavailability of Cd, Cu, Ni and Zn.
1.2.6 Rhizosphere processes
The rhizosphere is defined as the volume of soil affected by the presence of roots of growing plants, which is generally accepted as the area 1-2 mm from the root surface (Gregory, 2006). Biological associations with plant roots that may potentially affect plant availability of metals are mycorrhiza fungi, rhizobia (N-fixing bacteria associated with legumes), and other soil bacteria and fungi, whose roles are largely undefined owing to the vast number of bacterial species present in the rhizosphere. The biogeochemistry of Cd, Cu, Ni and Zn in the rhizosphere environment remains largely undescribed (Puschenreiter et al., 2005; Wenzel et al., 2003). The reason for this is that the rhizosphere involves many physical, chemical and biological components, which are intricately and intimately linked. Moreover the majority of rhizosphere research to date has mostly focused on plant nutrition and major nutrient (e.g. P and N) and C cycling in soil ecosystems (Gregory, 2006; Jones et al., 2004).
1.2.6.1 pH changes
Rhizosphere influence on the pH of the surrounding soil, is probably the most distinguishable and measureable chemical change that would influence the bioavailability of metals. The pH of the rhizosphere can vary by up to 2.5 pH units from that of the bulk soil solution, depending on the species of plant and buffering capacity of the soil + - (Marschner, 1995; Reichman, 2002). This is mostly due to the excretion of H /HCO 3 and to a lesser degree organic acids to balance the internal charge difference as a result of cation and anion uptake by roots, and also the production of CO2 from microbial activity (Jones et al., 2004; Marschner and Marschner, 1995). In addition, certain plants, referred
10 to as Strategy II plants (e.g. Zea mays), are known to acidify the surrounding media by production of phytosiderophores which are non-protein amino acids released from the roots to form soluble complexes with Fe3+ when experiencing deficiencies of Fe and also Mn (Marschner and Marschner, 1995; Reichman, 2002; Reid et al., 2003a).
1.2.6.2 Mycorrhiza and root exudates
Mycorrhizae are soil fungi symbiotically associated with the roots of most plant species. Mycorrhiza obtain photosynthates and in return increase the plant uptake of P and trace elements (Jones et al., 2004; Reichman, 2002). This increase in plant uptake of P and trace elements by mycorrhiza association is achieved by increasing the surface area from which nutrients are sourced (Reichman, 2002). Root exudates have been hypothesized to be involved in the enhanced mobilization and acquisition of many nutrients from the soil as well as the detoxification of metals. However, with few exceptions there is little mechanistic evidence from soil systems to support this theory of metal detoxification (Jones et al., 2004; Krpata et al., 2009; Zhao et al., 2001). There are many root exudates released into the rhizosphere, but for the most part their role is not clearly understood (Gregory, 2006). In P deficient soil, these exudates typically include low molecular weight organic anions such as citrate, oxalate, malate, lactate and fumarate (Jones et al., 2004). How these exudates affect metal bioavailability has been the subject of several studies (Cattani et al., 2006; Collins et al., 2003a; Collins et al., 2003b; Courchesne et al., 2008; Degryse et al., 2008; Lombnaes et al., 2008; Lu et al., 2007). These findings revealed changes in the speciation and solubility of Cd, Cu and Zn were significantly increased in the presence of organic anions such as citrate and oxalate. However, the extent to which these organo-metal complexes were being formed was affected by pH and the concentration of organic anions and of competing cations such as Ca (Collins et al., 2003b; Lombnaes et al., 2008; Lu et al., 2007). At present there is a dearth of information regarding the role of organometallic complexes in plant enhanced nutrient acquisition (Courchesne et al., 2008; Jones et al., 2004).
11
1.2.6.3 Rhizobia and microbial activity
In legumes a symbiotic association with rhizobia bacteria exists, that results in the
bacterial fixation of N2 from the atmosphere to supply the plant with accessible nitrogen (N). As a result legumes take up more cations that anions and acidify their rhizosphere as a consequence (Gregory, 2006). This modification of solution pH through N assimilation by rhizobia has in some cases lead to a reduction in the populations of these simbionts in metal contaminated soils (Chaudri et al., 2008; Kopittke et al., 2007). In another study by Cloutier-Hurteau et al. (2008), rhizosphere relationships between pH, microbial biomass N and Cu2+ also indicated that bacteria influenced Cu availability by modifying the pH of the solution via N assimilation.
Microbial mineralization of organic matter has been suggested to partly supply metals such as Cu, Ni and Zn to the solution fraction of the rhizosphere through root decay (Cloutier-Hurteau et al., 2008). Links have been established between urease activity, biomass variables, organic matter and DOC, and water soluble Cu that indicate the effect that bacteria have on Cu speciation and bioavailability (Cattani et al., 2006; Chaperon and Sauve, 2007; Cloutier-Hurteau et al., 2008; van Hees et al., 2004; Vogeler et al., 2008). However, the precise mechanisms and the degree of influence that bacteria have on the speciation of Cu in the rhizosphere still remain largely unknown (Cloutier-Hurteau et al., 2008).
The influence of bacteria in the rhizosphere of Ni accumulating plants and plants grown on serpentine soils has been described in a number of studies (El-Aziz et al., 1991; Idris et al., 2004; Idris et al., 2006; Ma et al., 2009). It was found that certain bacteria enhanced the plant availability of Ni by directly enhancing Ni accumulation in plant tissues from the production phytosiderophores which induced the dissolution of Ni bearing mineral, and indirectly by promoting shoot and root biomass (Ma et al., 2009; Wenzel et al., 2003).
12 1.3 Metals in plants
The response of plants to elevated metals concentrations has generated a substantial amount of literature and is driven by two primary questions (Palmer and Guerinot, 2009):
• can crops be grown safely and productively at elevated concentrations?
• how and why do certain species function successfully at these concentrations?
Thus, the following section describes various uptake and tolerance/toxicity mechanisms.
1.3.1 Mechanisms of metal uptake and transport
Cations, including micronutrients (i.e. Fe, Cu, Mn, Zn and Ni) and contaminants (i.e. Cd and Pb) are passively taken up through ion channels. The reason for this is that micronutrients and contaminants such as Cd are all charged to varying degrees hence membrane voltages will have a significant impact on their movement into the cells. Hence electrochemical gradients applying to the transport of a particular nutrient largely define the possible mechanisms for uptake (Reid et al., 2003a; Reid, 2001). Most plants have a membrane potential between -120 mV and -180 mV, and based on thermodynamics, major cation nutrients, such as Ca and Mg, and metal uptake is classified as passive (Reichman, 2002; Reid et al., 2003a).
Metal uptake occurs in small fluxes, usually as a result of small external solution concentrations and also their low internal requirements (Reid et al., 2003a). Because of the small fluxes of micronutrients across the membrane, it is difficult to ascertain if ion channels are solely responsible for uptake. As a result, there are few reliable data on trace element uptake into intact plant cells, and evidence for particular transport mechanisms remains largely indirect (Reid et al., 2003a).
Calcium ions are transported across the root membrane via non-selective channels and it is hypothesised that uptake (low affinity) of metals from typical soil solution concentrations may also occur via this pathway (Reid et al., 2003b). There are many examples of competition between trace metals for uptake (at the root surface) and also
13 between trace metals and macronutrient cations such as Ca and K (Heenan and Campbell, 1981; Kalis et al., 2006; Reid et al., 2003a; Warne et al., 2008b).
Examples of cation competition at the root surface suggest that non-selective channels facilitate the entry of both essential and non-essential metals. Hence competition for transport sites would favour the uptake of those cations at higher concentrations, thereby exacerbating nutrient deficiencies (Palmer and Guerinot, 2009; Reid et al., 2003a; Welch, 1995). Conversely, when supply of micronutrients is too low to meet the nutritional requirements of the plant there appears to be some molecular and physiological evidence for the induction of high-affinity trace metal transport systems in response to deficiency (Reid et al., 2003a). For example, in the case of induced Fe deficiency in Arabidopsis, protein expression (IRT1) resulted in the root membrane that also strongly stimulated the uptake of Mn and Zn, but not Cu (Connolly and Guerinot, 2002; Vert et al., 2002). Hence, Fe deficient plants can also accumulate high concentrations of Mn and Zn.
1.3.2 Toxicity responses to elevated metal concentrations in soils
Copper, Ni and Zn are essential elements for plant nutrition and development, yet the same properties that make these transition metals essential can also make them toxic in excess (Alloway, 2009; Broadley et al., 2007; Chaney and Oliver, 1996; Longhurst et al., 2004; Marschner and Marschner, 1995; Palmer and Guerinot, 2009). Copper and Ni are most often used for their redox properties, with Cu and Ni essential for the structure and catalytic function of the enzymes (Marschner and Marschner, 1995; Palmer and Guerinot, 2009). Zinc is primarily used for its ability to act as a lewis acid (Palmer and Guerinot, 2009). Moreover, the amount of metal taken up by plants varies significantly with species as well as marked differences in metal tolerance among species, which has been documented in numerous studies (Broadley et al., 2007; Gray and McLaren, 2005; Macnicol and Beckett, 1985; Marschner and Marschner, 1995; McLaughlin et al., 1996; Naidu et al., 2003; Reichman et al., 2006).
14 1.3.2.1 Cadmium Of all the heavy metals, Cd is recognised as posing the bigger threat to agricultural food quality (Chaney and Oliver, 1996; McLaughlin et al., 1996). Cadmium is recognised as the most mobile and bioavailable of the cationic trace metals and Cd has no known biological role in higher plants and therefore the presence of Cd is discussed in terms of tolerance and toxicity (Alloway, 1995a; Kiekens and Cottenie, 1985; Sloan et al., 1997). Most general symptoms of Cd toxicity in plants are stunted growth and chlorosis of younger leaves (An, 2004; Dheri et al., 2007) and a decrease in root elongation (Sun et al., 2005; Wang et al., 2009). However, concentrations of Cd in crop plants are likely to exceed food safety standards before phytotoxicity levels are attained (McLaughlin et al., 1996; McLaughlin et al., 2000).
1.3.2.2 Copper Interveinal foliar chlorosis is a common initial symptom of Cu toxicity (Lock et al., 2007; Reichman, 2002). With increasing exposure to excess Cu, leaf tips and margins of shoots can become necrotic and, in acute toxicity, leaves may become wilted before eventually becoming necrotic (Marschner and Marschner, 1995; Reichman, 2002). For most crop species, the critical toxicity level of Cu in the leaves is above 30 µg g-1 dry weight (Macnicol and Beckett, 1985). Excess Cu supply in the soil can also lead to root growth inhibition and in plants more sensitive to Cu, inhibition of root elongation and damage of the plasma membrane of root cells (Marschner and Marschner, 1995), and can also adversely affect the nodulation process of legumes (Kopittke et al., 2007). Even though high Cu concentrations can adversely affect shoot growth, typically it is root growth and form that is affected before any signs of above ground growth abnormality has been detected (Reichman, 2002).
1.3.2.3 Nickel Generally, the toxicity of Ni is almost always less than Cu, Zn and especially Cd (McLaughlin et al., 2000), and for crop plants critical toxicity levels are in the range of 10 µg g-1 to > 50 µg g-1 dry weight (Macnicol and Beckett, 1985; Marschner and
15 Marschner, 1995) for sensitive and moderately tolerant species. As with Cu, root growth and form is also affected in the presence of excess Ni in soil (Rooney et al., 2007b), and in sensitive species, root growth is severely inhibited at concentrations < 5 µM of Ni (Marschner and Marschner, 1995). The occurrence of Ni hyperaccumulators (e.g. genus Alyssum) and Ni tolerant plants growing on serpentine soils has been the focus of several studies as a result of their potential use in the phytoremediation of metal contaminated soils (Chaney et al., 2008; Chaney et al., 1997; Himmelbauer et al., 2005; Wenzel et al., 2003). In Ni hyperaccumulating species growing on serpentine soils, the leaves can reach 10– 30 mg g-1 dry weight, compared with 0.001 - 0.01 mg g-1 dry weight in most crop species (Marschner and Marschner, 1995). Tolerance in Ni hyperaccumulators is mostly achieved by complexation of Ni with organic ions such as malate and citrate (Homer et al., 1991).
1.3.2.4 Zinc Chlorosis of the younger leaves is typically the first symptom of Zn toxicity also with plants exhibiting Zn toxicity symptoms having smaller leaves than control plants (Reichman, 2002). With progressive exposure to toxic levels of Zn, this can develop into a reddening of the leaves and in severe cases lead to necrosis of the leaf tip (Marschner, 1995). In roots, the toxicity is manifested as a reduction in the growth of the main root, fewer and shorter lateral roots and a yellowing of the root (Broadley et al., 2007; Marschner and Marschner, 1995; Reichman, 2002). The critical toxicity levels in leaves of crop plants are from as low as 100 µg g-1 to more than 300 µg g-1, with the latter value being more typical (Marschner and Marschner, 1995). However, when the supply of plant available Zn is high, an induced deficiency of Fe and Mg can result because of the similar ion radius of Zn2+ and Fe2+ (Marschner and Marschner, 1995).
16 1.4 Methods for determining metal bioavailability in soils
Although bioavailability is recognized as the most relevant and critical aspect of metal fraction to measure, there is still remains no general consensus on how speciation and bioavailability can be best measured (Hooda, 2007; McLaughlin et al., 2000; Menzies et al., 2007; Zhang et al., 2001). Attempts to find suitable soil tests for estimating metal bioavailability have varied from measuring concentrations of the immediately bioavailable metal to also including measures of capacity of the soil to replenish the metal or buffer the metal concentration as it is removed from the soil solution (Reichman, 2002). This has resulted in the development of a multitude of soil tests of which, six methods currently used for estimating bioavailability will be discussed in this section.
1.4.1 Total extractable metals
The use of total metal concentrations in soil is still the most commonly used value to determine metal bioavailability and ecotoxicity (Menzies et al., 2007). Strong acids, such as HF, HClO4, HNO3 and aqua regia (HNO3/HCl) are commonly used to extract the ‘total’ metal in a soil (Blakemore et al., 1987). However, as discussed in section 2.2 many factors influence the bioavailability of metals to plants and organisms, making total metal a typically poor measure of bioavailability (McBride et al., 2003). However, there have been some good relationships documented between total metal and plant responses (Reichman, 2002) although these were soil specific, thereby demonstrating the use of total metal as a robust measure of bioavailability as ineffective (McLaughlin et al., 2000).
1.4.2 Chelation methods
Chelation techniques involving aminopolycarboxylic acids such a diethylene-triamine- penta-acetic acid (DPTA) and ethylene-diamine-tetra-acetic acid (EDTA) have been used
17 extensively to determine the phytoavailability of metals in soils (Hooda and Alloway, 1994b; McLaughlin et al., 2000; Meers et al., 2008; Reichman, 2002). These tests have been commonly used to determine metal deficiencies and have been shown to be good predictors of deficiencies (Reichman, 2002). Moreover, environmentally more acceptable chelating agents such as EDTA have also been assessed as a potential soil amendment for the remediation of metals in contaminated soils as an alternative to phytoextraction (Meers et al., 2008).
Although chelating compounds have been successfully used to determine metal deficiencies in soils, the use of these methods to determine phytotoxic concentrations of metals has been less so, with plant responses typically poorly correlated with DPTA or EDTA extractable metals (McLaughlin et al., 2000; Menzies et al., 2007; Reichman, 2002). There are a number of reasons proposed for this lack of strong correlation between plant and soil concentrations of metals: these methods were developed for use in soils containing low amounts of organic matter; the concentration of metals in the soils may exceed the chelating ability of the compound; the use of EDTA requires the compound to be pH buffered, typically pH 6 (McLaren et al., 1984) that would have a significant effect on metal solubility, particularly Cd and Zn (Menzies et al., 2007; Reichman, 2002).
1.4.3 Weak and neutral salt extractants
Weak and neutral salt extractants such as CaCl2, Ca(NO3)2 and NaNO3 have been evaluated in a number of recent studies (Datta and Young, 2005; Ettler et al., 2007; McBride et al., 2009; Meers et al., 2009; Menzies et al., 2007; Nolan et al., 2005) with promising results. The use of neutral salts as a method to predict bioavailability is based on the hypothesis that plant available metals are mostly sourced on soil mineral surfaces and can be easily displaced by other cations such as Ca and Na (Andrews et al., 1996; Houba et al., 2000; McBride et al., 2003; McLaughlin et al., 1999). In a review of extractants used to estimate metal availability, Menzies et al. (2007) found that neutral salt solutions tended to provide the best relationship between soil extractable metal and
18 plant tissue concentration. The review compared the relationships of six of these extractants (1 M NH4OAc and 1 M NH4NO3 for Cd, 1 M NaNO3 for Zn and 0.01 M 2 CaCl2 for Cd, Ni and Zn), and apart from NH4NO3 for Cd, all had r >0.5. Although good relationships were observed the use of neutral and weak salt extraction are relatively new, and consequently datasets are limited.
Of these extractants CaCl2 has been the most extensively reviewed, particularly in relation to assessment of plant available Cd (Ozkutlu et al., 2007; Weggler et al., 2004; Weggler-Beaton et al., 2000). The reason for this is that increased chloride concentrations has been shown to increase Cd concentration in soil solution and Cd uptake by plants (McLaughlin et al., 1994; Weggler et al., 2004). Studies comparing
CaCl2 extractable Cd and plant tissue concentrations have found reasonably successful correlations r2 >0.5 (Houba et al., 2000; Weggler et al., 2004) and these findings suggest that this type of extractant may also be suitable for metals such as Cu, Ni and Zn (Menzies et al., 2007). In addition, a study by Meers et al. (2009) to determine the effect - - of the counter ion (Cl vs NO3 ) found that at sufficiently high ionic strengths (e.g. 0.1 M)
Ca(NO3)2 can be used to give an estimate of the total exchangeable pool, whereas weaker extractions similar to the ionic strength of the soil solution (0.01 M) allows for a quantitative soil response of the metal released as a function of increasing Ca2+ in the extractant under relevant soil solution conditions.
1.4.4 Metals in soil solution (water extractable)
Soil solution is the fraction of metal considered immediately available for plant uptake (Gray and McLaren, 2006). Soil solution is defined as the aqueous liquid phase of the soil and its solutes (McBride, 1994). However, there appears no standard approach to the extraction and measuring of soil solution and methods of extraction range from centrifuging, to leaching, to extraction using Rhizon probes, and all of these techniques require differing ratios of soil to water (Degryse et al., 2009a; Reichman, 2002).
19 The amount of metal in soil solution and the ability of the soil to buffer metal concentrations in the soils have been described respectively as the intensity (I) and quantity (Q) factors (Tiller et al., 1972), and relationships between these, as the buffer power (C) (McLaughlin et al. 2000). The factors that affect (I) and (Q) in soil solution include the total concentration of metals in the soil, soil pH, cation exchange capacity, the nature of the exchange sites and the mass balance of metals and good correlations between soil solution Cd, Cu and Zn have been found with these soil properties (McBride et al., 1997a). McLaughlin et al. (2000) also provided a comprehensive review of factors affecting plant and soil biota metal availability through the modification of (I) and (Q).
Several investigations (Gray et al., 1999b; McBride et al., 1997a; Sauve et al., 2000a; Tambasco et al., 2000) have proposed that soluble metal concentrations in soils can be predicted using a semi-empirical equation. The theoretical background to this competitive adsorption model has been presented by McBride et al. (1997a), in which metal solubility in soils can be derived from a semi-empirical approach that assumes that the free metal (i.e. Cd2+) and H+ compete for adsorption on soil exchange sites. This suggests that these models may be useful for deriving certain metals loadings in soils and may also be used for risk assessment of contaminated sites. Even though some good relationships have been found with metal concentrations in solution and shoot concentration in some studies, these appeared dependent on soil type, plant species and metal, hence total metal in soil solution may not be a robust method for determining bioavailability (Reichman, 2002). The reason that total metal in soil solution may not be a good predictor of bioavailability is that not all forms of metal in solution are thought to be available for plants and soil organisms and it had been demonstrated that free ion and certain labile complexes of metals are available for uptake, while a significant portion of metal species in solution are relatively unavailable (Arnold et al., 2007; Broadley et al., 2007; Degryse et al., 2006a; 2006b; McLaughlin et al., 1997; Sauve, 2003; Slaveykova et al., 2009; Stacey et al., 2008).
20 1.4.5 Diffuse gradient in thin films
Previous extraction methods discussed also allow an insight into the geochemical forms of metals in the soil. However, these methods do not account for the ability of the soil to sustain the solution concentration following the depletion by plant uptake (Zhang and Davison, 2001). The diffusive gradient in thin film technique usually abbreviated to DGT was initially developed by Davison and Zhang, (1994) for the purpose of measuring in situ concentration and fluxes of trace element species in aqueous solutions (Degryse et al., 2003). It has since been modified to quantitatively measure the supply of labile trace metals in soils (Degryse et al., 2009b; Hooda et al., 1999; Zhang et al., 1995;1998; 2001).
The DGT device houses a resin-gel layer that is separated from the soil by a diffusion gel layer and a protective membrane. The technique relies on the accumulation of solutes onto the binding agent (such as the cation exchange resin) after passing through the uniform and well-defined diffusion gel layer. The binding agent acts as a localised sink for the solute thus causing a depletion of trace metals in the pore water and a re-supply from the labile metal pool on the solid phase, thereby introducing a diffusional flux to the DGT device from the surrounding solution; a process that simulates plant metal uptake (Degryse et al., 2003; Lehto et al., 2006a). The DGT-measured fluxes are affected by concentrations in the soil solution and the re-supply from the soil solid phase. This re- supply is dependent on the amount of labile metal sorbed onto the solid phase and the rate of exchange between the solid and the solution phase. This technique is unique in the respect that the resin-gel is selective only to labile ions.
In predicting metal uptake by plants, successful comparisons were made between kinetics of metal release in soils and prediction of the corresponding concentrations in plants using DGT (Zhang et al., 2004). Good correlations between metal concentrations in plants and metals have been measured by DGT in the same soil (Davison et al., 1999). Other studies demonstrated that under conditions where capacity and the kinetics of supply from the soil limit uptake, the controls on uptake to DGT and plant are similar (Degryse et al., 2009b; Nolan et al., 2005; Oporto et al., 2009; Zhang et al., 2004). The
21 use of DGT to predict metal uptake over a variety of plant species and soils types has been successful in some cases (Degryse et al., 2009b; Ernstberger et al., 2002; Nolan et al., 2005; Zhang et al., 2001). However, despite the growing use of this method to assess metal availability, particularly to plants, there is no standardised protocol for the preparation of soils prior to DGT deployment.
1.4.6 Speciation modeling and free ion activity
It is a widely held view that free metal ion activity is often a much better indicator of metal availability in soils than the total metal content (Cances et al., 2003; Degryse et al., 2009a; Lofts et al., 2004; Lofts et al., 2007; McBride et al., 1997a). Metals in soil solution may be present as free ions (e.g. Cd2+) or as complexes with inorganic anions - 2- 2- (e.g. Cl , CO3 , SO4 ), or organic anions (e.g. citrate and malate), or complexed with other forms of DOC. However, depending on the metal, the free ion may be the dominant form in solution and the free ion is generally the most reactive species in solution and thus can be a major determinant in bioavailability (Almas et al., 2006; Degryse et al., 2009a). Analytically quantifying the amount of free ion activity and labile species has only been attempted in a few studies (Sauve et al., 1995; Sauve et al., 2000b) and thus insufficient rigorous work has been done to confirm the hypothesis that free ion activity is the majority bioavailable species (Nolan et al., 2003; Reichman, 2002).
Assemblage models such as WHAM (Tipping et al., 2003b) consider the soil environment as a set of independent reactive surfaces and combines several models to describe sorption onto organic matter (soluble and solid), clays and oxides (Degryse et al., 2009a). These models require a considerable amount of information and several assumptions are made, about the percentage of active organic matter (typically 65 % fulvic acid in solution) and surface areas of the clays and oxides. Studies have shown that the complexation of Cu with DOC is less than predicted by WHAM (Buekers et al., 2008; Nolan et al., 2003; Sauve et al., 1997) and because few studies have described the complexation of Ni by DOC it is highly likely that the amount complexed may greatly differ than what would be predicted by WHAM (Degryse et al., 2009a; Nolan et al.,
22 2009). Generally only a few studies have compared measured solid and aqueous distribution of metals in soils with predicted values obtained from assemblage models (Amery et al., 2008; Buekers et al., 2008; Tipping et al., 2003b), and findings reveals that predictions and measured distributions differed considerable and, hence adjustments to the model were necessary. Studies that have compared predicted free ion activity with plant metal uptake found good relationships with Cd and Zn in spinach (Spinacia oleracea) and ryegrass (Lolium perenne) (Almas et al., 2006; Hough et al., 2005), and Ni2+ activity and toxic effects on barley roots (Rooney et al., 2007b) indicated that the free ion activities of Cd, Ni and Zn represent a significant proportion of the total bioavailable pool.
1.5 The use of biosolids as a soil amendment
Biosolids (sewage sludge) is the residual product of wastewater treatment and has been used widely in the agriculture industry for several years as it provides a valuable source of organic matter and macronutrients, as well as providing a means of disposing of an unwanted waste (Antoniadis and Alloway, 2001; McLaren et al., 2003; McLaughlin et al., 2007; Merrington and Smernik, 2004; Oliver et al., 2005; Smith, 2009; Speir et al., 2003b; Sullivan et al., 2006a). However, biosolids also contain organic and metal contaminants and there are some concerns regarding their environmental fate, and for this reason its agricultural use is regulated (McLaren et al., 2005; Speir et al., 2003a; 2003b; Sullivan et al., 2006b).
Two major environmental concerns arise from metal contaminants in biosolids following land application, namely the contamination of shallow groundwater from metal-rich leachate and the accumulation and uptake of metals in soils and plants (Ashworth and Alloway, 2004; Smith, 2009; Speir et al., 2003b). As a result of these environmental risks, rules and guidelines controlling both the quantity and quality of biosolids applied to land have been developed to assist in sustainable management and to provide advice on the safeguards necessary to protect public health (microbiological) and for metal contaminants. The focus on metal contamination is aimed primarily at preventing metal
23 accumulation in the food chain and toxic effects on plants and soil fauna (Alloway and Jackson, 1991; Antoniadis and Alloway, 2002a; 2003a; McLaren et al., 2007; McLaughlin et al., 2000; 2006; 2007; Speir et al., 2007).
Conversely, biosolids have been used extensively as soil amendments to reduce the bioavailability and phytotoxicity of metals in contaminated soils (Basta et al., 2001; Brown et al., 2005; Brown et al., 2007; Navas et al., 1999; Nwachukwu and Pulford, 2009; Stuczynski et al., 2007; van Herwijnen et al., 2007b). The hypothesis for the use of these soil amendments is that added organic matter may induce sorption or precipitation reactions for metals, resulting in increased metal retention on the solid phase and thus a reduction in the metal fraction available for plant uptake (Brown et al., 2005; Merrington et al., 2003; Nwachukwu and Pulford, 2009).
The efficacy of biosolids amendment at retaining metals has been examined by several studies (Antoniadis and Alloway, 2002b; 2003b; Antonladis and Tsadilas, 2007; Ashworth and Alloway, 2004; 2007; Brown et al., 2005; McBride et al., 1997b; McLaughlin et al., 2006; Merrington and Smernik, 2004; Merrington et al., 2003). The results from these studies have been varied with respect to the type of metal. Studies by Antoniadis and Alloway (2002b) and Ashworth and Alloway (2004) observed that biosolids possess relatively poor Ni retention capabilities. While Antoniadis and Alloway (2003a), van Herwijnen et al. (2007b) and Nwachukwu and Pulford (2009) observed an increase in the solubility of Ni and Zn overtime. Cadmium availability was not significantly affected, or availability decreased in the presence of biosolids (McLaughlin et al., 2006; Merrington and Smernik, 2004; Merrington et al., 1997). The efficacy of biosolids at reducing the bioavailability of metals is generally inconclusive but overall findings indicate potential risks in applying biosolids over longer time periods with respect to metal bioavailability and mobility.
24 1.5.1 Biosolids research and management in New Zealand
Metal pollution of agricultural land in New Zealand is concerning, although it is less severe than has been documented in many European countries (McLaughlin et al., 2000). However, New Zealand is economically dependent on agricultural products, which in- turn is dependent on maintaining adequate nutrient levels in soils. New Zealand environmental regulations and guidelines rely largely on data generated in northern hemisphere countries, which has led to unsuitable levels being chosen for several trace elements, especially Cd, Cu, Ni and Zn (Cavanagh and Coakley 2005; McLaughlin et al., 2000). As discussed in previous sections, the bioavailability of metals can be environmentally significant for several years after introduction (i.e. from the amendment of biosolids) and to predict metal bioavailability is the goal of ongoing research.
The majority of biosolids and soils research to date in New Zealand has been undertaken collaboratively by Environmental Science and Research Ltd. (ESR) and Lincoln University. The outcome of this research has helped to define guidelines for the safe and beneficial use of biosolids. Research to date has covered the trialing of five different New Zealand soils amended with biosolids and biosolids spiked with metal salts. These studies have assessed leachate quality, changes in soil chemistry plant metal concentrations and effects on soil organisms (Horswell et al., 2003; 2006; 2007; McLaren, 2005; McLaren et al., 2003; 2004; 2005; Yeates et al., 2006; Yuan, 2009) as well as impacts to groundwater quality (Speir et al., 2003b). Metal fractionation was examined in metal spiked biosolids and in leachate from high-rate application of biosolids to land (McLaren and Clucas, 2001).
Results from the metal spiked biosolids trials showed that after three years of continuous monitoring, soil pH had decreased and the concentrations of Cd, Ni and Zn in the drainage leachate had increased (McLaren et al., 2003; 2005). However, the lower pH had little effect on, or decreased the concentrations of Cr, Cu and Pb. Additionally, there were substantial differences observed between the soils, in the context of increased metal leachate, and Speir et al. (2003b) also showed that soil (coarse textured sand) could be strongly acidified by the sludge addition which resulted in considerable metal solubility,
25 mobility and bioavailability for plant uptake. A combination of plant cover, soil organic matter, clays and oxides contributed to the metal binding capacities of the soils and were responsible for the relatively low levels of leaching observed (McLaren et al., 2004). Speir et al. (2003b) confirmed that metals present in the biosolids used were largely unavailable as they were strongly complexed with organic matter contained in biosolids. However, it was observed under certain circumstances (low soil pH and clay content), metals contained in biosolids were more likely to move through the soil profile with the potential to contaminate groundwater (McLaren et al., 2004).
It is recognised that biosolids utilisation and management needs to be effectively managed in New Zealand and this has resulted in guidelines developed for the safe application of biosolids to land (NZWWA, 2003). These grade biosolids according to both stabilisation and contamination. Currently, the limits for metal contaminants in biosolids are given as total concentrations, which are set out in Table 1.1. As further research into determining the best measure of metal bioavailability continues, it is expected that guidelines will subsequently incorporate the available metal fraction with specified ecological receptors into land management practices (Cavanagh and Coakley 2005).
Table 1.1 Biosolids classification by contaminant levels grades a + b (Cavanagh and Coakley, 2005).
Contaminant Grade a (mg kg-1) Grade b (mg kg-1) As 20 30 Cd 1 10 Cr 600 1500 Cu 100 1250 Pb 300 300 Hg 1 7.5 Ni 60 135 Zn 300 1500
26 Chapter 2 Research Aim and Objectives
2.1 Aim
The overall aim of this thesis was to assess and validate measures of bioavailable Cd, Cu, Ni and Zn across contrasting soil types, ranges of metal concentration and presence of biosolids. This was accomplished by assessing six potential measures of bioavailability across a range of soil types spiked with metal salts in the presence and absence of biosolids using wheat (Triticum aestivum) and ryegrass (Lolium perenne) as biological response models.
2.2 Objectives
The overall aim of the research programme was fulfilled by the following three major objectives which were completed between 2006 and 2009:
1. Examining the effects of biosolids amendment on the bioavailability of Cd, Cu, Ni and Zn in contrasting soil types placed in controlled environmental conditions. This was achieved using wheat as the model plant system, where wheat seedlings were germinated in three soils spiked with increasing levels of metal salts which had been incubated for two and 24 weeks;
2. Examining the effects of biosolids amendment and metal salt applications on the bioavailability of Cd, Cu, Ni and Zn in contrasting soil types managed under field conditions over a 24-month period. This was achieved by using ryegrass as the model plant system, where ryegrass was grown in three soils spiked with increasing levels of metal salts, to which biosolids had been added to the lowest metal salt concentration treatment;
27 3. To assess the overall strength of the relationships determined between metal concentrations in two plant species and potential measures of Cd, Cu, Ni and Zn availability from a range of New Zealand soils amended with a combination of biosolids and metal salts. This was achieved by combining results in objectives 2 and 3 and data obtained from closely related biosolids and metal salt plot field studies carried out by Environmental Science and Research Ltd. (ESR) in collaboration with Lincoln University.
2.3 Structure
The previous chapter consisted of an introductory literature review that outlines mainstream concepts regarding bioavailability and methodologies used to predict bioavailability. Chapter 1 also briefly described plant mechanisms responsible for metal uptake and responses to metal toxicity and the use of biosolids as a soil amendment. Chapter 3 contains a list of soil descriptions and localities for each of the soil assessed in this study and similar studies carried out by ESR in collaboration with Lincoln University. Description of sample preparation and analytical methods common to studies described in subsequent chapters are also included.
The three main objectives are individually addressed and were the focus of Chapters 4, 5 and 6 respectively. Chapter 4 evaluated the effects of biosolids amendment on the bioavailability of Cd, Cu, Ni and Zn in soils maintained under controlled environmental conditions, using wheat as the biological response model; Chapter 5 investigated the effects of biosolids amendment and metal salt applications on the bioavailability of Cd, Cu, Ni and Zn in soils managed under field conditions using ryegrass as the biological response model; and Chapter 6 assessed the overall strength of the relationships determined between metal concentrations in wheat and ryegrass shoots, and measures of Cd, Cu, Ni and Zn bioavailability from a range of New Zealand soils amended with biosolids and metal salts. The findings from Chapters 4 – 6 were individually discussed at
28 the end of each chapter, while Chapter 7 concluded this thesis with a general discussion concluding remarks and approaches for further study.
29 Chapter 3 Soil Classification and Methodology
3.1 Soil localities and classifications
Three soils from the Canterbury region, a Halkett sandy loam (HK), Summit/Rapaki silt loam (SM) and Wakanui silt loam (WK) (Table 3.1) were selected for this study based on their contrasting soil properties (e.g. organic matter content, mineralogy and texture). The HK and WK soils were located in the wider Christchurch area, Canterbury (Figure 3.1), where the mean annual rainfall of the area is about 620 mm in the north-east and about 690 mm in the south-west (Cox, 1978). The HK soil series were formed on sand dunes and are located mostly within the north-west of the district. Wakanui soils were developed in fine greywacke alluvium brought in by ancient overflow channels of the Waimakariri River (Cox, 1978).
Summit soil was taken from the Port Hills area (elevation 275 to 300m a.s.l.) adjacent to the Canterbury Plains (Figure 3.1). These soils were derived from loess and basaltic material (Griffiths, 1974). The mean annual rainfall of the Port Hills area is approximately 1100 mm. The mean annual daily temperature for Canterbury is 11.0˚ C, day length is approximately 15.5 hours in mid summer (January) and 9 hours in mid winter (July) (Cox, 1978). Soils used in the joint Environmental Science and Research Ltd (ESR) and Lincoln University biosolids and metal salt trials are also included in Table 3.1. Biosolids for use in experiments described in Chapters 4 and 5 were collected in September 2006 from the Bromley wastewater treatment plant in Christchurch, New Zealand.
30
NORTH ISLAND
Hamilton (1 soil)
Taranaki region (1 soil)
Nelson (1 soil) Wellington (1 soil)
Christchurch
Canterbury region (6 soils)
SOUTH ISLAND
Figure 3.1 Localities of soils used in experiments described in Chapters 4, 5 and 6.
31 Table 3.1 Characteristics and locations of experimental soils.
Soil series NZ soil classification NZ Genetic Classification Location within Canterbury Halkett sandy loam Typic Sandy Recent Soil Yellow-brown sands West Melton Summit silt loam Orthic Brown Soil Upland Yellow brown earth Port Hills Wakanui silt loam Mottled Immature Pallic Gley soil – Yellow-grey Prebbleton, Ladbrooks Soil earth and Lincoln Soils used in Lincoln University and ESR trials Soil series NZ soil classification NZ Genetic Classification Location within New Zealand Ashley Dene deep Typic Fragic Pallic Soil Yellow-grey earth, weakly Canterbury fine sandy loam1 leached, moderately gleyed Egmont black Typic Orthic Allophanic Yellow-brown loam, Taranaki loam1 Soil moderately leached Foxton black sand1 Typic Sandy Brown Soil Yellow-brown sand Wellington Horotiu sandy Orthic Allophanic Soil Yellow-brown loam Hamilton loam4 Lismore silt loam3 Orthic Brown Soil Udic Ustochrept (USDA) Canterbury Tahunanui loamy Yellow Brown Sand Brown soil Nelson sand4 Otaki1 Weathered Fluvial Recent Recent soil Wellington Soil Templeton fine Immature Pallic Soil Intergrade between yellow- Canterbury sandy loam1,2,4 grey earth and recent soil Waihou silt loam1 Typic Orthic Pumice Soil Yellow-brown loam Hamilton
1 Lysimeter trials involving biosolids only, managed under field conditions 2 Pasture field trial involving biosolids spiked with metal salts 3 A forest soil field trial involving the application of biosolids spiked with metal salts 4 Metal salt plot trials managed under field conditions
32 3.2 Characterisation of soils and biosolids
All soils were air-dried at 20˚C before being ground and sieved through a 2 mm stainless steel sieve. Soils were then stored in polyethylene bags at room temperature prior to analyses.
Amorphous Al and Fe oxides: Air-dried soil (2.0 g) was mixed with 50 ml of 0.2 M oxalic acid and placed on an end-over-end shaker for 4 hours (in darkness). Samples were centrifuged at 2000 rpm for 10 minutes before being filtered through Whatman No. 42 filter paper (McKeague and Day, 1966). Iron and Al were analysed using an Inductively Coupled Plasma Optical Emissions Spectrophotometer (ICP-OES) Varian 720-ES fitted with a SPS-3 auto-sampler and ultrasonic nebuliser. Amorphous Al and Fe results are shown in Table 3.2.
Cation exchange capacity and exchangeable bases: The cation exchange capacity (CEC), base saturation and exchangeable bases of biosolids and soils was determined by ammonium acetate
leaching at pH 7.0 (Blakemore et al., 1987). Ca, Mg, Na, K and total NH4-N were analysed in the leachate using atomic absorption spectrophotometry (AAS) Shimadzu AA-670 (Shimadzu Oceania, Australia) and Flow Injection Analyser (FOSS FIAstar 5000 triple channel analyser, Foss Tecator, Hoganas, Sweden) (Table 3.2 andTable 3.3). Cation exchange capacity was calculated from NH4-N results. Total exchangeable bases (TEB) were the sum of Ca, Mg, Na and K. The percentage of base saturation is derived from the TEB/CEC.
Soil particle size distribution: Representative soil samples were sent to the University of Waikato New Zealand for texture analyses. Soil particle size distribution was determined using the Mastersizer 2000 (Table 3.4).
33 Soil pH: Approximately 10 g ± 0.2 of prepared soil was mixed with 25 ml of deionised water and left for six hours to equilibrate (Blakemore et al., 1987). Soil pH was recorded using a Mettler Toledo pH meter (Mettler Toledo, U.S.A.), (Table 3.2).
Total carbon: Total carbon was determined from 0.5 g of prepared soil (Gray et al., 1999a) placed in pre-weighed silica crucibles heated to 1300˚C by a LECO 2000 CNS Analyser (Leco, Australia) (Table 3.2).
Water holding capacity for soils: The maximum water holding capacity (MWHC) is a method used to determine the amount of water that a free draining soil can hold. Soil was dried (50 g) overnight at 105˚C and recording the weight. Soil was then placed in a small container fitted with fine-meshed gauze at the base, and allowing it to stand in a tray of water level with the top on the container. Soils were left for four hours before removing from the tray and re-weighing. The MWHC is the difference between the weight of the dry soil and the saturated soil and expressed as a percentage (Smart et al, 2004).
Table 3.2 Calculated MWHC, pH, Al and Fe oxides, CEC and organic carbon values for three New Zealand soils and biosolids sourced from the Bromley wastewater treatment plant used in experiments described in Chapter 4 and 5.
Soil Maximum pH Amorphous Al- Amorphous Cation Total water holding oxides Fe-oxides exchange carbon capacity (%) (mg kg-1) (mg kg-1) capacity (%) -1 (cmolCkg ) Halkett 56 5.6 95 176 8 3.1 Summit 67 5.6 525 696 22 7.4 Wakanui 65 6.1 144 313 16 3.5 Biosolids >100 8.3 na na 44 40.1 na = not analysed
34
Table 3.3 Calculated values of total base saturation (TBS), exchangeable bases (EB) and percentage of base saturation (BS) for three New Zealand soils and biosolids sourced from the Bromley wastewater treatment plant used in experiments described in Chapters 4 and 5.
-1 Soil TBS Exchangeable bases (cmolCkg ) Base saturation (%) (%) Ca Mg K Na Ca Mg K Na Halkett 67 4.5 0.34 0.21 0.07 59.0 4.5 2.7 0.9 Summit 38 4.3 2.58 1.07 0.24 19.7 11.9 4.9 1.1 Wakanui 68 7.7 1.37 1.39 0.15 49.3 8.8 8.9 1.0 Biosolids 65 16.5 8.31 1.97 1.56 37.6 19.0 4.5 3.5
Table 3.4 Results of the soil texture analysis determined for three New Zealand soils used in experiments described in Chapters 4 and 5.
Soil Particle size (µm) reported in volume %
0.06 – 2 2 – 63 63 - 2000 2000 Halkett 0.47 12.79 86.73 0
Summit 21.80 63.67 13.75 0
Wakanui 3.85 76.12 20.03 0
3.3 Plant and soil analyses using AAS, GFAAS and ICP-OES
All the metal concentrations in the extraction methods described below were analysed using a GBC Avanta Atomic Absorption Spectrophotometer (AAS), Graphite Furnace Atomic Absorption Spectrophotometer (GFAAS) GBC System 3000 and GBC 909AA with PAL 3000, or an Inductively Coupled Plasma Optical Emissions Spectrophotometer (ICP-OES) Varian 720- ES fitted with a SPS-3 auto-sampler and ultrasonic nebuliser.
35 3.3.1 Analysis of plant samples
Ryegrass and wheat from the seed and germination bioassays were prepared for analysis using the method described in Gray et al. (1999a) where 0.5 g of prepared plant sample was digested in
10 ml of 69% HNO3 (Aristar BDB grade) and heated on a digestion block (DK Heating 42 place digestion block, VELP, Italy) for 6 hours (up to 140° C). All plant samples were analysed for total Cd, Cu, Ni and Zn using ICP-OES. A reference standard (1573a Tomatoes Leaves, U.S. Department of Commerce, National Institute of Standards and Technology, Gaitherburg, MD 20899) was used during batches of analayses to check the accuracy of the results.
3.3.2 Analysis of soil samples for estimates of metal availability
3.3.2.1 Total recoverable metals
Total recoverable Cd, Cu, Ni and Zn was determined using methods described in Kovacs et al.
(2000), where 1 g of prepared soil is left overnight in 5 ml of 69% HNO3 (Aristar BDB grade).
The following day, 5 ml of 50% H2O2 was added to the pre-digested soils before placement on a 42 place digestion block (up to 120 C˚) for 4.5 hours. Samples were then filtered through a Whatman filter No. 42 filter paper and made up to 25 ml with deionised water. Samples from the first wheat seedling bioassay (Chapter 4) were analysed using AAS and GFAAS. The remainder of experimental samples were analysed by ICP-OES. A reference standard (2711 Montana Soil, Moderately Elevated Trace Element Concentrations, U.S. Department of Commerce, National Institute of Standards and Technology, Gaitherburg, MD 20899) was used during batches of analayses to check the accuracy of the results.
36 3.3.2.2 EDTA extractions
A solution of 0.04 M ethylene diamine tetra-acetic acid disodium salt (EDTA), adjusted to pH 6.0 with 0.1 M NaOH and was prepared according to McLaren et al. (1984). EDTA solution (25 ml) was added to 10 g of prepared soil, shaken for 2 hours then centrifuged at 15,000 rpm for 10 minutes. Samples were filtered through a Whatman No. 42 filter paper then analysed by AAS (first wheat seedling experiment, Chapter 4) and ICP-OES (remainder of experimental samples).
3.3.2.3 Calcium nitrate (CaNO3)2 extractions
A solution of 0.05M Ca(NO3)2 was prepared according to Gray et al. (1999a). Prepared
Ca(NO3)2 (30 ml) solution was added to 5 g of prepared soil, shaken for 2 hours then centrifuged at 15,000 rpm for 10 minutes. Samples were filtered through a Whatman No. 42 filter paper then analysed by ICP-OES.
3.3.2.4 Soil solution extractions
Soil solution was extracted from prepared lysimeter soils and soils used in the seed and germination bioassay experiments using Rhizon soil moisture samplers for soil solution extractions (MOM-type Rhizons; Eijkelkamp Agrisearch Equipment, Giesbeek, Netherlands). Soils were wetted to between 90 ± 10 % of their MWHC, left for 24 hours to equilibrate before being fitted with the soil moisture probes (Figure 3.2). The pH measurements of the solution extracted was measured and each sample was analysed for dissolved organic carbon (DOC), - - - - 3- 2- + anions (Br , Cl , NO3 , NO2 , PO4 , SO4 ), cations (Ca, K, Mg, Na, NH4 ) and trace metals (Cd, Cu, Ni, Zn). Total organic carbon was analysed using a Shimadzu Total Organic Carbon Analyser (TOC-5000A, Shimadzu, Australia), anions and cations were analysed using Dionex DX-120 Suppressed Ion Exchange Chromatograph (Dionex Corporation, California, U.S.A.),
37 Flow Injection Analyser (FOSS FIAstar 5000 triple channel analyser, Foss Tecator, Hoganas, Sweden) and ICP-OES. Cadmium, Cu, Ni and Zn were analysed using ICP-OES.
Figure 3.2 Soil solution extraction using Rhizon MOM type probes from soils used in the wheat seedling bioassays described in Chapters 4 and 6
38 3.3.2.5 Diffusive gradient in thin-films and DIFS analyses
Analysis by diffusive gradient in thin-films (DGT) was undertaken according to the methodology described by Zhang et al. (1998). Diffusive and chelex resin gels were made from a gel preparation solution, which consists of a 2% DGT gel cross-linker (DGT Research Ltd, Lancaster, United Kingdom) mixed with an acrylamide solution (40%) and double deionised water. For the diffusive gels the gel preparation solution was mixed with ammonium persulphate solution (10%), followed by a solution of N,N,N,N-Tetramethylethylenediamine (TEMED, 99%). The prepared diffusive gel solution was pipetted into acid washed glass plates held 0.8 mm apart. Diffusive gel moulds were then placed in an oven set at 45° C for 1 hour. Set diffusive gel moulds were then removed, washed in 1 L of MQ water, and then left to hydrate for 4 hours in a fresh 1 L beaker of MQ water. Hydrated diffusive gels were washed an additional two times with fresh MQ water before being cut into size (25 mm) and stored in 0.01 M NaNO3 solution until required.
The chelating resin was prepared using an ion exchange resin (Chelex-100), that had been pre- soaked in MQ water for 1 hour. Excess water was then decanted off the resin and 4 g of hydrated resin was added to the prepared gel solution and mixed thoroughly before a solution of ammonium persulphate solution (10%) was added followed by a solution of TEMED (99%). The resin gel mixture was then pipetted into acid washed glass plates held 0.4 apart and then placed in an oven set at 45° C for 1 hour. Set chelating gels were then removed and left to hydrate in 0.5 L of MQ water. Water was changed twice (every 4 hours), before gels were removed for cutting of moulds. Cut moulds (25 mm) were placed in 0.01 M NaNO3 solution until required.
The assembled DGT device consisted of a plastic backing plate, followed by a chelating resin gel (0.4 mm thick), a diffusive gel (0.8 mm thick) and a mixed cellulous ester membrane filter 0.45µm thick (Advantec, Japan) held in place by a plastic cap with a 17 mm window. Wetted soil (paste consistency) from each pot was smeared using a stainless steel spatula on the exposed filter window of the DGT device before placement onto the soils (Figure 3.2). DGT devices were left in the saturated soil for 24 hours. The resin gel layer was then removed from the DGT device and metals extracted into 10 ml 1M HNO3 prior to analysis using ICP-OES. The concentration at
39 the soil-diffusive layer interface (CDGT) and the effective concentration (CE) were calculated using equations 1-3:
The mass of the metals accumulated by the resin gel was calculated using Equation 1:
M = Ce(VHNO3 + Vgel)/fe Equation 1
where Ce is the metal concentration in the 1M HNO3, VHNO3 is the volume of HNO3, Vgel is the volume of gel and fe is the elution factor for metals.
The time averaged concentration at the soil-diffusive layer interface, CDGT, was calculated from Equation 2:
CDGT = M∆g/DtA Equation 2
where ∆g is the diffusion layer thickness, D is the diffusion coefficient of the metal in the diffusion layer at a given temperature, t is the deployment time (s) and A is the area in membrane exposed to the soil.
Effective concentration, (CE), was calculated from Equation 3:
CE = CDGT/RDIFF Equation 3
where RDIFF was calculated using the 2D-DIFS modelling programme (Harper et al., 2000) from soil density, porosity, and the diffusion coefficients for each of the metals in
water and soil. The value RDIFF represents the ratio of the mean interfacial concentration due to resupply by diffusion only to the initial or bulk solution concentration.
40
A B C
Figure 3.3 Washing of DGT device with de-ionised water after 24 hour placement on soil wetted to 90 ± 10% MWHC and allowed to equilibrate. Photo B shows removal of filter and diffusive gel, and photo C shows placement of resin gel in 10 ml of 0.01 M HNO3.
3.3.2.6 Speciation and solid-solution phase modelling using WHAM 6.0
Soil solution speciation and solid-solution partitioning of Cd, Cu, Ni and Zn was done using WHAM (Windermere Humic Aqueous Model) version 6.0 (Tipping, 1994; 1998). The WHAM version 6 modelling programme is a discrete site and electrostatic model for simulating cation binding to humic substances. Electrostatic accumulations of ions adjacent to the molecular surfaces (non-specific binding), is simulated with Donnan-type expressions. Complexation reactions with inorganic ligands are also accounted for. WHAM 6.0 also includes a surface complexation model (SCM) with parameters for ion-binding to oxides of aluminium, iron. Manganese and silicon, and a model for cation exchange on clay.
Input parameters into the model for soil solution speciation were soil solution concentrations of - - - - 3- 2- Cd, Cu, Ni and Zn, major anions, (Br , Cl , NO3 , NO2 , PO4 , SO4 ) and cations (Ca, K, Mg, Na, + NH4 ). Soil solution pH, temperature (assigned value of 278 K), pCO2 (assigned value of 0.0003 atm) were also required. Fulvic acid in solution was assumed to be 65% (Tipping et al., 2002) of the calculated amount of DOC.
41 Solid phase input parameters for modelling solid phase and soil solution partitioning consisted of - 3- total soil concentrations of Cd, Cu, Ni and Zn; major anions (Cl , Nitrate-N, Nitrite-N, PO4 , 2- 2+ + 2+ + + SO4 ), and major cations (Ca , K , Mg , Na , NH4 ), temperature, fulvic acid and pH as determined from soil solution extractions. Clay content was determined from textural analyses (Table 3.4), Al and Fe oxides using the oxalic acid extract methods (Table 3.2), particulate humic acid and fulvic acid calculated from the total soil carbon content (Table 3.2), which is assumed to be a 50:50 distribution of both, humic acid and fulvic acid (Stevenson et al., 1984; Stevenson, 1994).
42 Chapter 4 Evaluating the Effects of Biosolids Amendment on the Bioavailability of Cd, Cu, Ni and Zn in Soils Maintained Under Controlled Environmental Conditions
4.1 Introduction
Incubation of soils and germination of plants under controlled environmental conditions have been commonly used to investigate soil process dynamics, efficacy of soil amendments (e.g. biosolids), and changes in chemistry of metal contaminated soils (Amery et al., 2007; Brown et al., 2005; Buekers et al., 2008; Dussault et al., 2008; Öncel et al., 2000; Rooney et al., 2007a).
Biosolids and composts have been used extensively as soil amendments to reduce the bioavailability and phytotoxicity of metals in contaminated soils (Nwachukwu and Pulford, 2009; van Herwijnen et al., 2007b). The hypothesis for the use of these soil amendments is that added organic matter may induce sorption or precipitation reactions for metals, resulting in increased metal retention on the solid phase and thus a reduction in the metal fraction available for plant uptake (Nwachukwu and Pulford, 2009). The success of organic amendments is dependent upon the type of soil and source of the organic matter (Brown et al., 2005; Navas et al., 1999; Ruttens et al., 2006; Santibanez et al., 2008; van Herwijnen et al., 2007a; 2007b).
In some instances, a positive environmental effect of reduced metal availability resulting from organic amendments can also coincide with an unexpected effect of increased metal solubility. Amending soils with biosolids provides additional dissolved organic carbon (DOC) to the system. The presence of DOC, particularly humic acids, can mobilise soil nutrients and metals via decreased pH and chelation by soluble organic ligands as well as reduce the availability of metals by complexation (Antoniadis and Alloway, 2002b; Khan et al., 2006).
Increases of metal bioavailability as a result of additional DOC can occur in a variety of ways, including extra sorption sites provided by biosolids may be lost rapidly as a result of microbial
43 decomposition (Hooda and Alloway, 1994a; McBride, 1995); the acidic groups contained in DOC such as carboxyl and phenolic OH functional groups decrease pH, thereby increasing metal solubility (Khan et al., 2006; Marschner and Kalbitz, 2003) and HA competing for free metal ions to form soluble complexes, thereby reducing metal adsorption onto soil surfaces and retaining these bioavailable organo-metal complexes in soil solution (Antoniadis and Alloway, 2002b; Ashworth and Alloway, 2004). Thus, while it is often hypothesized that organic-rich amendments will reduce metal bioavailability this is not always the case (Antoniadis and Alloway, 2002b; van Herwijnen et al., 2007a; van Herwijnen et al., 2007b). Therefore, further work is needed to elucidate the impact of biosolids, on heavy metal bioavailability in agricultural environments.
Obtaining an effective method to measure metal bioavailability in soils has been the subject of much research and is still inconclusive (Hamon et al., 1997; Hooda, 2007; Zhang and Young, 2005; Zhang et al., 2001). In particular, studies investigating the efficacy of metal bioavailability predictions have not yet established the effect of organic amendments on the validity of assays for metal bioavailability in soils. Therefore, the experiments in this study were used to determine the effects of organic amendments on the validity of metal bioavailability assays.
The main objective of this study was to examine the effects of biosolids amendment on the bioavailability of Cd, Cu, Ni and Zn in contrasting soil types placed in controlled environmental conditions. This was achieved using wheat (Triticum aestivum) as the model plant system. Plants were grown in three soils spiked with increasing levels of metal salts which had been incubated for two and 24 weeks.
Metal bioavailability was determined by comparing shoot metal concentrations with six different assays of soil metal extractability. Results of metal bioavailability assessments were compared to establish which assay was most successful across soils and amendments. The data from this study was used to assist in the validation of bioavailability measures to predict plant concentrations of Cd, Cu, Ni and Zn in a diverse range of New Zealand soil types, which is presented in Chapter 6.
44 4.2 Methods
4.2.1 Set-up of incubation trials
Approximately 90 kg each (three 25 L volume buckets) of Halkett (HK), Summit (SM) and Wakanui (WK) soils (Table 3.1) were collected from the top 10 cm of each soil profile. Each soil was divided equally into 90 portions immediately after collection, with each portion weighing 3.0 ± 0.1 kg, (three replicates per treatment and 10 treatments per soil). Soil portions were placed in 2 L plastic containers, and each portion received a single application of combined increasing concentrations of Cd, Cu, Ni and Zn or one of Cd only (Table 4.1). A treatment of Cd only was included as the considerably higher concentrations of Cu, Ni and Zn added may mask any effects that could result from the addition of Cd.
Metals were added as hydrated metal sulphates. Biosolids, as received, were sourced from the Christchurch City Council treatment works at Bromley and added to half of the metal treated soils at a rate equivalent to 400 kg N ha-1 (fresh biosolids collected for this study comprised on average 5.6% of N). The rate of biosolids application was based on the maximum rate of application to agricultural soils as outlined in the guidelines regarding the safe application of biosolids to land (NZWWA, 2003). Cadmium and Ni concentrations were also based on current biosolids guideline levels (NZWWA, 2003). Copper and Zn concentrations were chosen based on previous complimentary work by Environmental Science and Research (Ltd) and the Centre for Soil and Environmental Quality, Lincoln University (McLaren and Clucas, 2001; Speir et al., 2007).
Fresh treated soil portions were watered with de-ionised water to field capacity. Soils were then placed in an incubator at 25 ± 2˚C in darkness and mixed at two week intervals. Subsamples of the treated soils were removed from the incubators after two weeks and 24 weeks for use in the seedling bioassay experiments.
45 Table 4.1 Soil treatments for incubation experiments. Metal were added in the form of hydrated sulphates and biosolids were applied at a rate equivalent to 400 kg N ha-1.
Treatment Abbreviations Cd added Cu added Ni added Zn added (mg kg-1) (mg kg-1) (mg kg-1) (mg kg-1)
Control C - - - -
Biosolids B - - - - Biosolids + BLM 1 30 50 70 low metals Low metals LM 1 30 50 70 Biosolids + BMM 5 200 150 300 medium metals Medium MM 5 200 150 300 metals Biosolids + BHM 10 750 300 1000 high metals
High metals HM 10 750 300 1000
Biosolids + BCd 10 - - - Cd only
Cd only Cd 10 - - -
4.2.2 Seedling bioassay experiments
Seedling emergence experiments were undertaken according to the methods in Smart et al. (2004). Full descriptions of plant and soil sample preparation and analytical methods were provided in section 3.3.
Before use in the seedling bioassay, samples of incubated treated soils were air dried at 23°C, ground then sieved through a 2 mm stainless steel sieve. The dried and sieved soil was weighed 200 ± 5 g into 90, 250 ml plastic pots and wetted to 50% of the soils maximum water holding capacity (MWHC), and left to equilibrate for 24 hours. Twenty-two seeds of bread wheat (Triticum aestivum sourced from Luisetti Seed Grain Merchant, Rangiora, New Zealand) were
46 sown into each pot of soil. The pots were then placed in controlled light (PAR of 400 µmol m-2 s- 1) growth cabinets at 21 C˚ (day) and 15 C˚ (night) on a 12 hour light/dark cycle for 20 days. After five days, cotyledons visible above the soil surface were counted as having emerged. On the 5th day emerged seedlings were then thinned to 10 plants per pot and 5 ml of nutrient solution (Ruakura solution, Smart, et al. (2004)) was added on days five, 10 and 15 after thinning. Pots were weighed daily and watered with deionised water as necessary to ensure that soils were maintained at 50% MWHC for the duration of the experiment. On day 20, seedlings were removed from the growth cabinets and photographed to record any visual differences in treatment responses. Seedlings were then harvested, washed thoroughly using deionised water and fresh weights recorded for each pot. Wheat shoots were placed overnight in a drying oven at 45 ± 1 °C, removed and weighed to record dry weight per pot. Dried plants were cut with stainless steel scissors into lengths of <0.5cm before being acid digested using 10 ml of concentrated HNO3 per sample and placed on a heating block cycle (section 3.3.1). Digested plant samples were then analysed for total Cd, Cu, Ni and Zn (section 3.3).
Extraction of soil solution and diffusive gradient in thin films (DGT) placement (sections 3.3.2.4 and 3.3.2.5) were undertaken simultaneously on the remaining soil in the pots after plant harvest.
Soils were also analysed for total, EDTA extractable, and Ca(NO3)2 extractable metal concentrations (sections 3.3.2.2 and 3.3.2.3). Soil pH and soil solution pH were also measured (section 3.2). Concentrations of DOC, major anions and cations in the soil solution were also measured (section 3.3.2). Soil extraction samples were analysed for total Cd, Cu, Ni and Zn using ICP-OES, AA and GFAAS (section 3.3.2). Salinity of the soil solution was defined as the - + 2- sum of Cl , Na and SO4 concentrations (Taiz et al., 1998).
Speciation and free ion activity of metals in soil solution were determined using WHAM 6.0.
(Tipping, 1998). Effective concentration (CE) was calculated from metal fluxes measured using DGT devices placed on soils and the supplementary modelling freeware 2D DIFS (Harper et al., 2000; Sochaczewski et al., 2007). Full calculations and descriptions of DGT placement, and speciation modelling using WHAM 6.0., were given in sections 3.3.2.5 and 3.3.2.6.
47 4.2.3 Derivation of critical levels of metals in plant tissues
Critical levels (CL) of Cd, Cu, Ni and Zn in wheat shoot tissue were obtained from Macnicol and Beckett (1985) and used to compare against plant results. Critical levels are based on a statistically significant 10 % reduction in dry weight yield.
4.2.4 Statistical analyses
Comparisons of plant metal concentrations between biosolid and non-biosolid amended soils, incubation times and treatment concentrations were analysed using paired t-tests in Sigma Plot 8.02 (SPSS, UK). Linear regressions for shoot metal concentrations and soil measured of bioavailability also used Sigma Plot 8.02. Correlations between variables and differences between metal treated soils amended and unamended with biosolids were determined using grouped and separate linear regression analyses performed in GenStat 11 (VSN International, UK).
4.3 Results
4.3.1 Changes in soil chemistry
4.3.1.1 Major cations, dissolved organic carbon and pH
Comparisons between biosolids amended and unamended metal spiked soils revealed significant increases in concentrations of dissolved organic carbon (DOC), soil solution salinity, and Ca and Mg ions in soil solution as a result of the addition of biosolids, irrespective of metal treatment concentrations (Table 4.2). Overall, the increased length of soil incubation time resulted in a significant decrease in soil and soil solution pH, as well as an increase in salinity levels of the soil solution after 24 weeks (Table 4.3).
48
Table 4.2 Significance levels for comparisons of general soil parameters between biosolids amended and unamended metal spiked soils (n = 180). Differences were considered significant at P<0.05 and highlighted in bold.
Variable P-value Soil pH 0.768 Soil solution pH 0.196 DOC (mg L-1) 0.001 Salinity of soil solution (mg L-1) 0.002 Soil solution Ca concentration (mg L-1) 0.027 Soil solution Mg concentration (mg L-1) 0.001
Table 4.3 Significance levels for comparisons of soil parameters between all treated soils sampled after two weeks and 24 weeks of incubation, n = 90. Differences were considered significant at P<0.05 and highlighted in bold.
Variable P-value Soil pH < 0.001 Soil solution pH 0.009 DOC (mg L-1) 0.24 Salinity of soil solution (mg L-1) < 0.001 Soil solution Ca concentration (mg L-1) 0.65 Soil solution Mg concentration (mg L-1) 0.35
4.3.1.2 Metal extractability
Results from pooled soil data showed that increased metal rate resulted in an increase in metal extractability (Appendix A). The presence of biosolids significantly increased the total amount of Cu and Zn in the soils (Table 4.4). With the exception of DGT results for Zn, which showed a significant decrease in effective concentration (CE), the presence of biosolids had no significant effect on EDTA extractable, Ca(NO3)2 extractable or soil solution metal concentrations, or on
49 DGT measured metal solubility and free ion activity (Table 4.4). In contrast, the longer soil incubation time of 24 weeks resulted in significant increases in the availability of Cd, Ni and Zn but not Cu compared to an incubation period of two weeks (Table 4.5). These results show that biosolids amendment did not alter the bioavailability of these metals as measured by five of the assays, irrespective of soils or treatment concentrations, but that the bioavailability of these metals did significantly increase with soil incubation time.
Table 4.4 Significance values for comparisons of six potential measures of metal bioavailability between biosolids amended and unamended metal spiked soils (log transformed data, n = 180). Differences were considered significant at P<0.05 and highlighted in bold.
Method Cd Cu Ni Zn Total metal 0.090 <0.001 0.060 0.002 EDTA 0.190 0.940 0.818 0.433
Ca(NO3)2 0.760 0.700 0.409 0.415 Soil solution 0.320 0.853 0.289 0.855 DGT 0.320 0.344 0.667 0.033 Free ion activity 0.320 0.335 0.785 0.816
Table 4.5 Significance levels for comparisons of six potential assays of metal bioavailability between treated soils incubated for two weeks and soils incubated for 24 weeks (log transformed data, n = 90). *Total metal was only measured on one occasion (24 weeks incubated soils). Differences were considered significant at P<0.05 and highlighted in bold.
Method Cd Cu Ni Zn Total metal * * * * EDTA 0.98 0.97 0.78 0.22
Ca(NO3)2 0.28 0.95 0.64 0.008 Soil solution 0.002 0.128 0.005 0.002 DGT <0.001 0.113 0.006 0.001 Free ion activity 0.001 0.98 0.003 0.001
50 4.3.2 Responses of plants
Over the course of the experiment, the typical changes in appearance of the seedlings were increased chlorosis, discolouration and relative stunting of plants with increasing metal treatment concentrations. These symptoms of phytotoxicity also increased in severity over the length of the experiment. Plants grown in the control pots showed no signs of discolouration or necrosis for the duration of the experiment (Figure 4.1). However, chlorosis and necrosis of some leaf tips were seen in plants grown in two of the soils (HK and SM). There were no observable changes in plant appearances for treatments B, BLM and LM across all soils. Plants grown in biosolids and medium metal treatment began to show signs of leaf discolouration (Figure 4.2). Signs of reduced shoot size first appeared in BMM treated HK soil (Figure 4.2) with symptoms of stunting, chlorosis, discolouration and necrosis were most severe in plants grown in treatments BHM and HM (Figure 4.3).
Figure 4.1 A. Appearance of wheat seedlings after 20 days of growth in untreated (control) soil previously incubated for six months at 50% of maximum water holding capacity. Scale is 10 centimetres.
51 A Biosolids + medium metals
HK SOIL SM SOIL WK SOIL B
HK SOIL
Figure 4.2 A. Appearance of wheat seedlings after 20 days of growth in all three soils treated with a combination of biosolids and 5, 200, 150 and 300 mg kg-1 of Cd, Cu, Ni and Zn (biosolids and medium metals), and incubated for 24 weeks at field capacity. B. Close-up of seedlings grown in HK soil showing first signs of chlorosis in some of the leaves as a result of soil treatment. Scale is 10 centimetres.
52 A Biosolids + high metals
HK SOIL SM SOIL WK SOIL B High metals
HK SOIL SM SOIL WK SOIL
Figure 4.3 A. Appearance of wheat seedlings after 20 days of growth in all three soils treated with a combination of biosolids and 10, 750, 300 and 1000 mg kg-1 of Cd, Cu, Ni and Zn, and incubated for 24 weeks at field capacity. B. Appearance of seedlings in all three soils grown in 10, 750, 300 and 1000 mg kg-1 of Cd, Cu, Ni and Zn without the addition of biosolids. Scale is 10 centimetres.
53 4.3.2.1 Estimations of metal bioavailability
Log-log relationships between pooled shoot metal concentrations and the six measures of metal extractability from pooled soil results were significant (P < 0.001) for all four metals.
Cadmium: Both EDTA and Ca(NO3)2 extractable Cd gave the strongest correlations of the six extraction methods with shoot Cd concentrations (r2 = 0.63 and 0.62 respectively, Figure 4.4B -1 and C). The spread of data for Ca(NO3)2 concentrations lower than 0.01 mg kg was most likely indicative of natural attenuation in the soils (Figure 4.4C). A distinctive clustering of data for soil solution results was observed which is most likely attributed to the Cd only treatments (Figure 4.4D)
Copper: Even though, all six extraction methods gave significant relationships with shoot Cu concentrations, (Figure 4.5) the strength of the correlations for each were poor with the highest recorded correlation obtained for total Cu (r2 = 0.15, Figure 4.5A) and the lowest obtained for soil solution Cu (r2 = 0.06, Figure 4.5D)
Nickel: Calcium nitrate extractable Ni gave the strongest correlation with shoot Ni concentrations (r2 = 0.73, Figure 4.6C) with soil solution Ni and Ni2+ activity yielding the next strongest and very similar correlations (r2 = 0.66 and 0.67 respectively, Figure 4.6D and F).
Zinc: As with Cd and Ni, Ca(NO3)2 extractable Zn yielded the strongest correlation of the six extraction methods with shoot Zn concentrations (r2 = 0.55, Figure 4.7C). In comparison, soil solution Zn and Zn2+ gave the poorest correlations (r2 = 0.33 for both, Figure 4.7D and F).
54 100 100 (A) (B)
10 10
1 1
0.1 0.1
y = -0.14 + 1.09x, r ²= 0.60, p< 0.001 y = 0.67 + 0.15x, r ²= 0.63, p< 0.001 0.01 0.01 0.1 1 10 100 0.01 0.1 1 10 100 -1 -1 Log10 (Total Cd, mg kg ) Log10 (EDTA-Cd, mg kg )
100 100 (C) (D) ) -1 10 10
1 1
0.1 0.1
(shoot concentration, mg kg mg concentration, (shoot y = 0.48 + 0.43x, r ²= 0.62, p< 0.001 y = 0.22 + 0.74x, r ² = 0.30, p< 0.001 0.01 0.01 10 10-5 10-4 10-3 10-2 10-1 100 101 10-9 10-8 10-7 10-6 10-5 10-4 10-3 10-2 10-1 100 101 -1 Log Log (Ca(NO ) -Cd, mg kg ) -1 10 3 2 Log10 (Soil solution-Cd, mg L )
100 100 (E) (F)
10 10
1 1
0.1 0.1
y = 0.42 + 0.52x, r ²= 0.25, p< 0.001 y = 0.24 + 2.16x, r ² = 0.30, p< 0.001 0.01 0.01 0.001 0.01 0.1 1 10 100 10-13 10-12 10-11 10-10 10-9 10-8 10-7 10-6 10-5 10-4 -1 Log (C - Cd, mmol L ) 2+ -1 10 E Log10 (Cd activity, mol L ) Figure 4.4 Relationships determined between shoot Cd concentrations and Cd concentrations in soil obtained using various methods: (A) Total Cd, (B) EDTA-extractable Cd, (C) Ca(NO3)2 extractable Cd, (D) soil solution, (E) effective concentration, (F) free ion activity. A linear regression line is fitted for each plot. n = 175.
55 1000 1000 (A)(A) (B)(B)
100 100
10 10
1 1
0.1 0.1
y = 0.39 + 0.27x, r ² = 0.15, p< 0.001 y = 0.62 + 0.17x, r ² = 0.13, p< 0.001 0.01 0.01 1 10 100 1000 0.01 0.1 1 10 100 1000 -1 Log (EDTA-Cu, mg kg-1) Log10 (Total Cu, mg kg ) 10
1000 1000 ) (C) (D)
-1 (D)
100 100
10 10
1 1
0.1 0.1 (shoot concentration, mg kg 10 y = 0.90 + 0.16x, r ² = 0.10, p< 0.001 y = 0.95 + 0.12x, r ²= 0.06, p< 0.001
Log 0.01 0.01 0.001 0.01 0.1 1 10 100 0.0001 0.001 0.01 0.1 1 10 100
-1 -1 Log10 (Ca(NO3)2,-Cu, mg kg ) Log10 (soil solution Cu, mg L )
1000 1000 (E)(E) (F)(F)
100 100
10 10
1 1
0.1 0.1
y = 0.82 + 0.16x, r ² = 0.13, p< 0.0001 y = 1.34 + 0.11x, r ² = 0.11, p< 0.001 0.01 0.01 0.01 0.1 1 10 100 1000 10-16 10-15 10-14 10-13 10-12 10-11 10-10 10-9 10-8 10-7 10-6 10-5 10-4 -1 Log (Cu2+ activity, mol L-1) Log10 (CE - Cu, mmol L ) 10
Figure 4.5 Relationships determined between shoot Cu concentrations and Cu concentrations in soil obtained using various methods: (A) Total Cu, (B) EDTA-extractable Cu, (C) Ca(NO3)2 extractable Cu, (D) soil solution, (E) effective concentration, (F) free ion activity. A linear regression line is fitted for each plot. n = 175.
56 10000 10000 (A) (B)
1000 1000
100 100
10 10
1 1
0.1 0.1
y = -2.03 + 1.88x, r ² = 0.63, p< 0.001 y = 0.05 + 0.78x, r ²= 0.60, p< 0.001 0.01 0.01 1 10 100 1000 0.01 0.1 1 10 100 1000 -1 -1 Log10 (EDTA-Ni, mg kg ) Log10 (Total Ni, mg kg )
10000 10000.00 ) (C) (D) -1
1000 1000.00
100 100.00
10 10.00
1 1.00
0.1
(shoot concentration, mg kg 0.10
10 y = 0.33 + 0.72x, r ²= 0.73, p< 0.001 y = 1.01 + 0.67x, r ²= 0.66, p< 0.001 0.01 Log 0.01 0.001 0.01 0.1 1 10 100 1000 0.0001 0.001 0.01 0.1 1 10 100 -1 -1 Log10 (Ca(NO3)2-Ni, mg kg ) Log10 (soil solution Ni, mg L )
10000 10000 (E) (F) 1000 1000 100 100 10
10 1
0.1 1
0.01 0.1 0.001 y = 0.37 + 0.57x, r ² = 0.44, p< 0.001 y = 3.91 + 0.58x, r ²= 0.67, p< 0.001 0.01 0.0001 0.01 0.1 1 10 100 1000 10000 10-12 10-11 10-10 10-9 10-8 10-7 10-6 10-5 10-4 10-3 -1 2+ -1 Log (C Ni, mmol L ) Log10 (Ni activity, mol L ) 10 E Figure 4.6 Relationships determined between shoot Ni concentrations and Ni concentrations in soils obtained using various methods: (A) Total Ni, (B) EDTA-extractable Ni, (C) Ca(NO3)2 extractable Ni, (D) soil solution , (E) effective concentration, (F) free ion activity. A linear regression line is fitted for each plot. n = 175.
57 10000 10000 (A) (B)
1000 1000
100 100
10 10
y = 3.91 + 0.58x, r ² = 0.38, p< 0.001 y = 1.32 + 0.42x, r ²= 0.54, p< 0.001 1 1 10 100 1000 0.1 1 10 100 1000 -1 -1 Log10 (Total Zn, mg kg ) Log10 (EDTA-Zn, mg kg )
10000 10000 )
-1 (C) (D)
1000 1000
100 100
10 10 (shoot concentration, mg kg 10
Log y = 1.58 + 0.36x, r ²= 0.55, p< 0.001 y = 1.91 + 0.26x, r ²= 0.33, p< 0.001 1 1 0.001 0.01 0.1 1 10 100 1000 0.001 0.01 0.1 1 10 100 1000 -1 -1 Log (soil solution Zn, mg L ) Log10 (Ca(NO3)2 Zn, mg kg ) 10
10000 10000 (E) (F)
1000 1000
100 100
10 10
y = 1.58 + 0.36x, r ²= 0.43, p< 0.001 y = 2.94 + 0.18x, r ² = 0.33, p< 0.001 1 1 0.01 0.1 1 10 100 1000 10-11 10-10 10-9 10-8 10-7 10-6 10-5 10-4 10-3 10-2 -1 Log (C Zn, mmol L ) 2+ -1 10 E Log10 (Zn activity, mol L ) Figure 4.7 Relationships determined between shoot Zn concentrations and concentrations of Zn in the soil
obtained using various methods: (A) Total Zn, (B) EDTA-extractable Zn, (C) Ca(NO3)2 extractable Zn, (D) soil solution, (E) effective concentration, (F) free ion activity. A linear regression line is fitted for each plot. n = 175.
58 Comparisons of the coefficients of determinations for all assays of bioavailability and shoot
metal concentrations, revealed that Ca(NO3)2 extractable metals gave the strongest relationship for shoot and soil data overall, closely followed by EDTA extractable metals. Cadmium, Cu and Zn concentrations in the soil solution gave the poorest correlations with corresponding shoot metal concentrations. However, for Ni the relationship determined between soil solution Ni and shoot Ni concentrations was second highest and similar to the r2 value obtained for Ni2+ activity
and shoot Ni concentrations (Figure 4.6). Relationship strengths for CE were similar and strongest for Ni and Zn and shoot concentrations (Figure 4.6E and 4.7E). However these r2 values were considerably lower than Ca(NO3)2 extractable and EDTA extractable correlations (Figure 4.6 B-C and 4.7B-C).
4.3.2.2 Dry matter production
Dry matter production significantly increased (P< 0.001) as soil incubation time increased (Figure 4.8). Dry weight appeared to decrease with increasing metal concentrations but the comparison of pooled treatment results for biosolids amended metal spiked soils gave no significant differences in dry weight (P = 0.08) due to the presence of biosolids. However, further analysis of the relationship between dry weight and Ca(NO3)2 extractable Cd, Cu, Ni and Zn concentrations revealed that Cu and Ni significantly decreased dry matter production irrespective of the presence of biosolids, while Cd and Zn only significantly decreased dry matter production in the presence of biosolids (Figure 4.9). Correlations were strongest for relationships between dry weight and extractable metal concentrations in metal treated soils amended with biosolids (Figure 4.9)
59 0.8 Key two weeks C = Control 24 weeks B = Biosolids BLM = Biosolids + low metals 0.6 LM = Low metals BMM = Biosolids + medium metals MM = Medium metals BHM = Biosolids + high metals 0.4 BCd = Biosolids + Cd Cd = Cd only Dry weight (g)
0.2
0.0 C B BLM LM BMM MM BHM HM BCd Cd
Treatment
Figure 4.8 Mean dry weights of wheat seedlings grown in biosolids amended and unamended metal treated soils incubated for two weeks and 24 weeks at field capacity. Bars denote the SEM, n = 89 for both trials.
60 (A) (B) 1.2 1.2 Cd Cu 1.0 1.0
0.8 0.8
0.6 0.6
0.4 0.4 Dry weight (g) Dry weight (g) 0.2 0.2
0.0 0.0 y = 0.56 + -0.09x, r2 = 0.20, p < 0.001 (B + MS) y = 0.41 + -0.14x, r ² = 0.27, p < 0.001 (B + MS) y = 0.41 + -8.5x, r2 = 3.56, p = 0.618 (MS) y = 0.37 + -0.05x, r ² = 0.07, p = 0.035 (MS) 0123456 0.001 0.01 0.1 1 10 100 -1 Ca(NO ) extractable Cd concentration (mg kg ) -1 3 2 Log10 (Ca(NO3)2 extractable Cu concentration, mg kg ) (C) (D) 1.2 1.2 Ni Zn 1.0 1.0
0.8 0.8
0.6 0.6
0.4 0.4 Dry weight(g) Dry Dry weight (g) 0.2 0.2
0.0 y = 0.49 + -0.10x, r ² = 0.17, p < 0.001 (B + MS) 0.0 y = 0.58 + -0.13x, r ² = 0.19, p < 0.001 (B + MS) y = 0.42 + -0.04x, r ² = 0.06, p = 0.0257 (MS) y = 0.46 + -0.05x, r ² = 0.05, p = 0.058 (MS)
0.01 0.1 1 10 100 1000 0.1 1 10 100 1000
-1 -1 Log (Ca(NO ) extractable Ni concentrations, mg kg ) Log (Ca(NO ) extractable Zn concentration, mg kg ) 10 3 2 10 3 2
Biosolids + metal salts (B + MS) Fitted line (B + MS) Metal salts (MS) Fitted line (MS)
Figure 4.9 A-D. Relationships determined between plant dry weight and corresponding Ca(NO3)2 extractable metal concentrations in biosolids amended and unamended metal treated soils, (A) = Cd, (B) = Cu, (C) = Ni, (D) = Zn, n = 179.
Other soil chemistry variables, such as pH, major cations and the salinity of the soil solution, were explored for significant correlations and of these it was found that only dry weight production correlated significantly with salinity (P< 0.001). Furthermore, comparisons with correlation strengths of metals and dry weights, and salinity and dry weights, revealed that salinity gave the highest correlation for pooled soil data (r2 = 0.28) and that every 100 mg L-1
61 increase in salinity resulted in a 5% reduction in dry weight production (Figure 4.10). Further examination of the relationship between salinity and dry weight revealed that the addition of biosolids significantly decreased the overall dry weight production (P = 0.017) compared to the unamended metal spike soils.
1.2 y = 0.61 + -7.57x, r ² = 0.33, p < 0.001 (B + MS) Biosolids + metal salts (B + MS) y = 0.51 + -4.87x, r ² = 0.19, p < 0.001 (MS) Fitted line (B + MS) 1.0 Metal salts (MS) Fitted line (MS)
0.8
0.6
0.4 Dry weightDry (g)
0.2
0.0
0 100 200 300 400 500 600 700
-1 Salinity (mg L ) Figure 4.10 Relationship determined between plant dry weight and the salinity of the soil solution, n = 179.
These results indicate that while the addition of biosolids did not significantly alter the effect that each of the four metals had on dry weight production (P = 0.426, 0.171, 0.231 and 0.189 for Cd, Cu, Ni and Zn respectively), it did result in significant change in the gradients for Cd and Cu, i.e. biosolids altered the relationship between these metals and dry weight (Cd, P = 0.004 and for Cu, P = 0.030). Although the strongest influence exerted on dry weight production resulted from an increase in amounts of salinity as a consequence of biosolids amendment rather than additional metal loading from either biosolids or metal salt application (Table 4.2).
62 4.3.2.3 Metal concentrations in wheat shoots
Soil incubation time had a significant effect (P < 0.001) on wheat shoot metal concentration and generally metal concentrations were lower in plants grown in soils incubated for 24 weeks (Table 4.6). There were also significant differences (P < 0.05) in shoot metal concentrations between soil types (Appendix A).
Cadmium: The critical level (CL) for wheat of 43 mg kg-1 for Cd was not reached at either sampling time (Figure 4.11) and incubation time had no significant effect on overall shoot Cd concentrations (P = 0.12). Comparisons of shoot Cd concentrations in plants grown in biosolids amended metal spiked soils to unamended metal spiked soils revealed a significant decrease in shoot Cd concentrations in plants grown in biosolids amended soils for both soil incubation times (Table 4.6). This significant decrease in shoot Cd concentration in the presence of biosolids was observed across all levels of metal treatment concentrations (Table 4.7)
Copper: Concentrations of Cu in the shoots were significantly higher in plants grown in soils incubated for two weeks (P< 0.001) compared with plants grown in soils incubated for 24 weeks. The CL for wheat of 17 mg kg-1 for Cu was only significantly (P<0.05) exceeded in plants grown in the two week incubated BHM and HM treated soils (Figure 4.12). The presence of biosolids resulted in significant increases in shoot Cu concentrations in plants gown in soils incubated for two weeks (Table 4.6). However, this significant effect was not observed in plants grown in soil incubated for 24 weeks (Table 4.6). Moreover, comparisons of shoot Cu concentrations between biosolids amended and unamended metal spiked soil at each treatment level revealed no significant differences (P>0.05, Table 4.7).
Nickel: Nickel concentrations in the shoots generally increased with increasing treatment concentrations and the CL for wheat of 46 mg kg-1 was exceeded in treatments HM and BHM for both soil incubation periods (Figure 4.13). Overall soil incubation time resulted in no significant differences in shoot Ni concentrations (P = 0.90) and amendment with biosolids did not produce a significant effect on shoot Ni concentrations at either soil incubation period (Table 4.6). However, at the lowest level of metal treatment concentration the presence of biosolids did
63 significantly increase shoot Ni concentrations (Table 4.7), although this increase was relatively small (Figure 4.13).
Zinc: Plants grown in soils incubated for two weeks had significantly higher shoot Zn concentrations (P = 0.009) compared with plants grown in soils incubated for 24 weeks. The CL for shoot Zn concentrations in wheat (224 mg kg-1) was exceeded in the highest metal treatment at both soil incubation periods (Figure 4.14). The addition of biosolids to metal spiked soils did not significantly impact on overall shoot concentrations (Table 4.6). Furthermore, comparisons at individual metal treatment levels revealed no significant differences in Zn shoot concentrations (Table 4.7).
Key two weeks C = Control 24 weeks B = Biosolids
) BLM = Biosolids + low metals -1 CL LM = Low metals BMM = Biosolids + medium metals MM = Medium metals BHM = Biosolids + high metals 40 BCd = Biosolids + Cd 20 Cd = Cd only
10 Cd concentration (mg kg
0 C B BLM LM BMM MM BHM HM BCd Cd Treatment Figure 4.11 Mean Cd concentrations in wheat shoot tissue (mg kg-1 dry weight) grown in biosolids amended and unamended metal spiked soils incubated for two weeks and 24 weeks. The critical level (CL) for Cd as depicted by the red line is 43 mg kg-1 (Macnicol and Beckett, 1985). Bars denote the SEM, n = 89 for both trials.
64 50 Key two weeks C = Control 24 weeks B = Biosolids
) 40 BLM = Biosolids + low metals -1 LM = Low metals BMM = Biosolids + medium metals 30 MM = Medium metals BHM = Biosolids + high metals BCd = Biosolids + Cd Cd = Cd only 20 CL
Cu concentration (mg kg 10
0 C B BLM LM BMM MM BHM HM BCd Cd Treatment Figure 4.12 Mean Cu concentrations in wheat shoots (mg kg-1 dry weight) grown in biosolids amended and unamended metal spiked soils incubated for 2 weeks and 24 weeks. The critical levels (CL) for Cu as depicted by the red line, is 17 mg kg-1 (Macnicol and Beckett, 1985). Bars denote the SEM, n = 89 for both trials.
65 250 Key two weeks C = Control 24 weeks B = Biosolids 200 ) BLM = Biosolids + low metals -1 LM = Low metals BMM = Biosolids + medium metals 150 MM = Medium metals BHM = Biosolids + high metals BCd = Biosolids + Cd 100 Cd = Cd only
Ni concentration (mg kg concentration Ni CL 50
0 C B BLM LM BMM MM BHM HM BCd Cd Treatment Figure 4.13 Mean Ni concentrations in wheat shoots (mg kg-1 dry weight) grown in biosolids amended and unamended metal spiked soils incubated for 2 weeks and 24 weeks. The critical levels (CL) for Ni as depicted by the red line, is 46 mg kg-1 (Macnicol and Beckett, 1985). Bars denote the SEM, n = 89 for both trials.
700 Key two weeks C = Control 600 24 weeks B = Biosolids
) BLM = Biosolids + low metals -1 500 LM = Low metals BMM = Biosolids + medium metals MM = Medium metals 400 BHM = Biosolids + high metals BCd = Biosolids + Cd 300 Cd = Cd only CL 200 Zn concentration (mg kg (mg concentration Zn
100
0 C B BLM LM BMM MM BHM HM BCd Cd
Treatment Figure 4.14 Mean Zn concentrations in wheat shoot tissue (mg kg-1 dry weight) grown in biosolids amended and unamended metal spiked soils incubated for 2 weeks and 24 weeks. The critical levels (CL) for Zn as depicted by the red line, is 224 mg kg-1 (Macnicol and Beckett, 1985). Bars denote the SEM, n = 89 for both trials.
66 Table 4.6 (A) Significance levels for comparisons between shoot metal concentrations in plants grown in biosolids-amended and unamended metal spiked soils (log transformed data, n = 35). (B) Significance levels for comparisons between two and 24 weeks incubation (log transformed data, n = 35). Differences were considered significant at P<0.05 and highlighted in bold.
(A) Comparisons between biosolids amended and unamended metal spiked soils Cd Cu Ni Zn Two weeks 0.003 0.024 0.390 0.388 incubation 24 weeks incubation 0.042 0.368 0.449 0.415 (B) Comparisons between soil incubation times (two weeks vs 24 weeks) Biosolids amended 0.229 <0.001 0.877 0.008 metal spiked soils Unamended metal 0.695 <0.001 0.112 0.013 spiked soils
Table 4.7 Significance levels for comparisons between shoot metal concentrations in plants grown in biosolids-amended and unamended metal treated soils, in order of increasing metal treatment levels (log transformed data, n = 70). Differences were considered significant at P<0.05 and highlighted in bold.
Metal treatment Cd Cu Ni Zn concentrations BLM vs LM <0.001 0.062 <0.001 0.234 BMM vs MM <0.001 0.062 0.954 0.234 BHM vs HM <0.001 0.062 0.954 0.234 BCd vs Cd 0.001 - - -
67 4.3.3 Effects of biosolids amendments on relationships between soil extraction methods and metal concentrations in shoots
The assessment of relationship strengths between six different extraction methods of metals from soils and total metal concentrations in plants (Figure 4.4 – 4.7) revealed
that Ca(NO3)2 extractable metals gave the best estimate of bioavailability. Therefore,
Ca(NO3)2 extractable metal concentrations were used to explore any potential impacts that amendment of metal spiked soils with biosolids may have on correlation strengths of bioavailability estimates.
There were no significance differences in relationships determined between extractable metal concentrations and shoot metal concentrations in the presence or absence of biosolids for Cd, Cu, Ni or Zn (Figure 4.15). A linear regression analysis of the grouped data revealed that for Cd, the addition of biosolids significantly reduced overall shoot Cd concentrations (P = 0.023, Figure 4.15A). This significant result for the reduction of Cd concentration in shoots in the presence of biosolids also supports findings presented in section 4.3.2.3 (Table 4.6 and 4.7). Comparisons made of the gradients of Cd bioavailability estimates between biosolids amended soils and unamended soils revealed no significant changes (P = 0.625), indicating that the
relationship between Cd concentrations in shoots and Ca(NO3)2 extractable Cd in the presence or absence of biosolids were not significantly different. Consequently this
result provides evidence that Cd bioavailability as measured by Ca(NO3)2 is not dependent on the presence or absence of biosolids.
For Cu, Ni and Zn, the addition of biosolids did not significantly alter Cu, Ni or Zn shoot concentrations (P= 0.389, P = 0.576, P = 0.598 for Cu, Ni and Zn respectively). These findings are also supported by results presented in section 4.3.5.3 (Table 4.6 and 4.7). Furthermore, there were no significant differences found between the gradients of biosolids amended soils and unamended soils for either Cu (Figure 4.15B), Ni (Figure 4.15C) or Zn (Figure 4.15D). These results are consistent with the
findings for Cd, that Cu, Ni and Zn bioavailability as measured by Ca(NO3)2 is not dependent on substrate type.
68 (A) (B) 100 120 Cd Cu y = 0.39 + 0.74x, r ² = 0.49, p< 0.001 (B+MS) y = 13.43 + 6.53x, r ²= 0.13, p< 0.001 (B+MS) 100 y = 0.52 + 0.68x, r ²= 0.50, p< 0.001 (MS) y = 10.77 + 4.19x, r ² 0.19, p< 0.001 (MS)
80 10
60
40
1
) 20 -1
0
0.1 -20 0.01 0.1 1 10 0.001 0.01 0.1 1 10 100 (C) (D) 10000 10000
concentration, mg kg mg concentration, Ni Zn y = 0.18 + 0.87x, r ²= 0.78, p< 0.001 (B+MS) 1000 y = 1.45 + 0.44x, r ²= 0.57, p< 0.001 (B+MS) y = 0.24 + 0.83x, r ²= 0.80, p< 0.001 (MS) y = 1.54 + 0.43x, r ²= 0.59, p< 0.001 (MS) (shoot 1000 10 100 Log
10 100
1
10
0.1
0.01 1 0.01 0.1 1 10 100 0.1 1 10 100 1000
-1 Log10 (Ca(NO3)2 extractable metal, mg kg )
Biosolids and metal salts Fitted line (biosolids and metal salts) Metal salts Fitted line (metal salts)
Figure 4.15 Metal bioavailability estimates as determined by Ca(NO3)2 extractions as affected by biosolids amendment of three soils spiked with metal salts: (A) results for Cd; (B) results for Cu; (C) results for Ni; (D) results for Zn .n = 71 (Biosolids + MS) and n = 72 (MS).
Previously described results for the CE of Zn in biosolids amended soils and unamended soils revealed significant difference between the groups (Table 4.4).
Hence, the relationship between concentrations of Zn in shoots and CE were compared for biosolids amended and unamended metal spiked soils (Figure 4.16). For both biosolids amended and unamended metal spiked soils there was a significant relationship (P< 0.001). Comparisons made between the gradients for biosolids amended and unamended soils yielded no significant difference (P = 0.423). These
69 findings confirm that the presence of biosolids did not alter the relationship between metal concentrations in shoots and CE-Zn and thus, providing further validation that soil estimates of metal bioavailability are independent of biosolids amendments.
10000
) Zn Biosolids + metal salts (B + MS) -1 Fitted line (B + MS) Metal salts (MS) Fitted line (MS) 1000
100
10 (shoot Zn concentration, mg kg concentration, Zn (shoot
10 y = 1.74 + 0.30x, r ² = 0.38, p <0.001 (B + MS)
Log y = 1.79 + 0.35, r ² = 0.47, p < 0.001 (MS) 1 0.01 0.1 1 10 100 1000
-1 Log10 (Effective concentration, mmol L )
Figure 4.16 Relationship between shoot Zn concentration and effective concentration as affected by amendment with biosolids, n = 71 (Biosolids + MS) and n = 72 (MS).
4.4 Discussion
4.4.1 Influence of biosolids on the availability of metals in soils
The addition of biosolids significantly (Table 4.2 and 4.4) increased the concentration of DOC, salinity, Ca and Mg in soil solution as well as increasing the total amount of soil Cu and Zn. These differences resulted from the higher concentrations of these parameters present in biosolids than levels found naturally in the soils used in this study (section 3.2, Appendix A). The amendment of metal spiked soils with biosolids did not significantly affect metal availability as measured by the six methods assessed (Table 4.4). The partitioning of metals in the biosolids amended and unamended metal
70 spiked soils were also compared and these results concurred with the findings of the soil extractability methods (Appendix A).
While addition of biosolids did not alter the extractability of these metals, soil incubation time did significantly increase the availability of Cd, Ni and Zn, which may be attributed to the significant (Table 4.3 and 4.5) decrease in soil and soil solution pH that was also observed. Traditionally, soil organic matter decomposition is considered one of the main contributing factors to decrease in soil pH (Helyar and Porter, 1989). This along with microbial and chemical degradation of biosolids would most likely contribute to the decrease in pH as well as the production of significantly higher levels of DOC (Table 4.2). However the degradation of biosolids alone did not contribute to the decrease in pH (Table 4.2). Thus a combination of soil organic matter decomposition, biosolids decomposition and adsorption to soil particulate matter that results in release of H+ are the mostly likely factors that resulted in the decrease in pH.
The decrease in pH after 24 weeks of incubation is likely to have been the main factor contributing to the increase in dissolved Cd, Ni and Zn concentrations (Table 4.5) and several studies have linked the solubility of Cd, Ni and Zn to the pH of the soil (Gray and McLaren, 2006; McBride et al., 1997a; Sauve et al., 2000a; Stephan et al., 2008) Another possible cause for increased concentrations of dissolved Cd, Ni and Zn, is the presence of increased amounts of DOC. Corresponding increases in the concentration of these metals in the solution phase and DOC has been found in a number of studies (Amery et al., 2007; Antonladis et al., 2008; Ashworth and Alloway, 2004; Doig and Liber, 2007; McBride et al.; 1997b; van Herwijnen et al., 2007a). Overall DOC concentrations did not change with increased incubation time, although the presence of biosolids did result in an increase in DOC compared to unamended soils. The pH of the soil and soil solution decreased with increased incubation time, while biosolids amendment of soils did not result in any pH changes. Increased Cd, Ni and Zn in solution with incubation time cannot directly be attributed to DOC it must therefore be attributed mainly to the change in pH.
Interestingly, the percentage of Cu and Ni present in the dissolved phase as modelled by WHAM 6.0. were predicted to be higher in the soil incubated for 24 weeks than soil incubated for two weeks. Actual measurements of Cu and in particular Ni,
71 directly measured after two and 24 weeks of incubation revealed that Cu and Ni were present in lower concentrations after 24 weeks of incubation compared with concentrations measured after two week of incubation (Appendix A). Although no direct explanation for the modelled and measured discrepancies can be deduced from this study, two recent studies by Amery et al. (2007; 2008) demonstrated that the mobilising potential of Cu depends upon the degree of humification of dissolved organic matter (DOM). This finding may also partly explain the difference in measured and modelled Ni (Amery et al., 2007). Nickel is similar to Cu in that it is has strong affinity for binding to organic matter (Sauve, et al., 2000), and so the degree of aromaticy may also contribute to the increased mobilisation of Ni as no relationship between total DOC and an increase in dissolved Ni concentrations was observed.
Unlike Cu and Ni, Cd affinity for DOC is rather small as Cd adsorbs weakly to organic matter, clays and oxides unless the pH is greater than 6 (Degryse et al., 2007). As a majority of the pH values were below 6, DOC would have had little influence in the solubility of Cd. While incubation time clearly had an effect on the solubility of Cd (Table 4.5), the presence of biosolids did not (Table 4.4). Thus it appears that the solubility of Cd is more strongly influenced by pH than DOC, and this relationship has been demonstrated by many studies (Antoniadis and Alloway, 2002b; Collins et al., 2003b; Degryse et al., 2007; Gray and McLaren, 2006; McBride et al., 1997a Meers et al.; 2005; Shi et al., 2007). Likewise, the same relationship with pH and solubility can be described for Zn (Antoniadis et al., 2008).
4.4.2 Responses of plants grown in metal spiked soils in the presence and absence of biosolids 4.4.1.1 Dry weight reduction
Examination of relationships between soil variables revealed two factors controlling the production of dry matter, namely salinity of the soil solution and, concentrations of Cu and Ni in the shoots (Figure 4.9 and 4.10). Of these two factors, salinity had the strongest relationship with dry weight decreases (Figure 4.10). While increased incubation time resulted in an increase in salinity (Table 4.3), salinity was
72 significantly higher in soils amended with biosolids irrespective of the incubation period (Table 4.2). The appearance and texture of plants grown in soils containing both medium to high levels of metals and biosolids, suggested that plants were under severe stress either as result of toxic concentrations of metals or osmotic pressures, or both (Figure 4.2 and 4.3). While it was expected that degrees of phytotoxicity would be exhibited by plants grown in the medium to high metal treatments, the osmotic stress induced by the presence of biosolids was unforeseen. Biosolids used in this study contained higher concentrations of exchangeable Na, K, Ca and Mg than were measured in the unamended soils used in this study (Table 3.3). In addition, metal salt dissolution would most likely have contributed to the overall increase in salinity in all amended soils. While dry matter production was most strongly influenced by salinity levels in the current study, the concentrations measured in the soil solution were well below levels considered saline (40 mM of NaCl which generates an osmotic pressure of approximately 0.2 Mpa) for soils (Munns and Tester, 2008).
Concentrations of Cu, Ni and Zn were present above critical thresholds in plants grown in the highest metal treatments, irrespective of the presence of biosolids (Figure 4.12 and 4.14). While Cu, Ni and Zn concentrations in shoots exceeded the CL in one or both incubated soil periods, only Cu and Ni gave significant correlations with dry weight reduction in the absence of biosolids (Figure 4.9B and C). The morphological symptoms of interveinal discolouration and leaf chlorosis, observed in the plants were highly suggestive of metal toxicity (Figure 4.2 and 4.3), where the appearance of interveinal foliar chlorosis is commonly linked to initial stages of Cu toxicity (Reichman, 2002), and interveinal chlorosis of leaves is often ascribed to Ni toxicity (Molas and Baran, 2004). However, further analysis would be needed to quantify and separate the effects that Cu or Ni had on plant growth.
4.4.1.2 Metal concentrations in shoots
Overall, Cd concentrations in shoots were reduced in plants grown in metal spiked soils amended with biosolids compared to unamended metal spiked soils. This outcome is in disagreement with the non-significant changes found between Cd soil extractability in biosolids amended and unamended soils (Table 4.4). Increased DOC,
73 salinity, Ca and Mg were the only significant changes in soil chemistry observed between amended and unamended metal spiked soils. Correlations between these variables and Cd concentrations in the shoots were only significant for DOC, but this relationship was poor (P = 0.019, r2 = 0.06, Appendix A) and salinity, which was also poorly correlated (P< 0.001, r2 = 0.19, Appendix A). Thus no strong direct relationship could be inferred from these soil chemistry parameters and reduced plant Cd uptake. While findings in this study appear to support the retention of Cd in biosolids amended soils, limitations in the experimental set-up and analyses make it difficult to accurately attribute the mechanism(s) by which Cd retention is occurring. Studies on the availability and uptake of Cd in plants have identified pH and DOC as the major influential factors controlling soil availability (Collins et al., 2003b; Gray et al., 1999b; 2003; McLaughlin et al., 2006). As pH did not significantly change with the addition of biosolids, this can not be directly attributed to the reduction in Cd concentrations in shoots measured in this study. From these results two possible mechanisms may be responsible for this effect:
• Dissolved organic matter sourced from biosolids reducing the availability of Cd in soil solution to the plants via complexation of free Cd2+ and weakly sorbed Cd (Merrington and Smernik, 2004);
• Addition of contaminants and salts in biosolids (i.e. Zn and Ca ions) which competitively inhibit the uptake of Cd by plants. This mechanism assumes that metal uptake by plants is mediated by a generic, divalent cation transporter and, many studies have confirmed that this is the most likely transport mechanism for micronutrient uptake (Reid et al., 2003a).
The addition of biosolids did not result in significant changes in Cu, Ni (with the exception of the low treatment) and Zn concentrations in the shoots compared to unamended treatments (Table 4.6 and 4.7), which is in agreement with the non- significant changes found in the soil extractability/solubility of these metals (Table 4.4). However, increased soil incubation time did result in significant reductions in Cu and Zn concentrations in the shoots (Table 4.6). The reduction of Cu and Zn concentration in shoots may be attributed to increased complexation of free metal ions by DOC, although estimations of speciation distributions between the two soil incubation periods do not clearly correspond to this hypothesis (Appendix A).
74 Complexation of metals by DOC could partly account for the decrease in shoot concentrations Cu and Zn, although competitive uptake as result of increased salinity in the soil solution with increased incubation time (Table 4.2) could also explain the reduction in Cu and Zn shoot concentrations. Further work would be required to elucidate the mechanism(s) responsible for the reduction in Cu and Zn shoot concentrations and the relationship with soil incubation time.
4.3.2 Estimates of metal bioavailability
Pooled soil and plant data revealed that Ca(NO3)2 extracted concentrations of Cd, Ni and Zn gave the strongest relationship with plant shoot concentrations (Figure 4.4 – 4.7). Correlation strengths for ETDA and total metal concentrations were generally
second highest, while soil solution, CE and free ion activity were often the weakest
correlations. Of the four metals, Ca(NO3)2 extractable Ni and corresponding shoot concentrations revealed the highest correlation (r2 = 0.73). Nickel results for the remaining five extraction methods were also the most closely correlated (Figure 4.6). Results for Ni indicate that plant uptake of Ni is strongly indicative of the amount of Ni extracted from the soil using the six methods assessed, particularly 0.05 M 0f
Ca(NO3)2. Hence the solubility of Ni as measured by Ca(NO3)2 extraction can reliably predict the degree of phytotoxicity and growth limitations resulting from excess amounts present in soils. The robustness of this method is further validated by the fact that the correlation was obtained across a range of soils, metal concentrations as in the presence and absence of biosolids.
Coefficient of determination values between extractable amounts of Cu from the soil and shoot Cu concentrations were poor in comparison to Cd, Ni and Zn. Copper is an essential trace element whose concentration is strongly regulated by the plant and is known to accumulate in, or around the roots, thus correlation between shoot Cu concentrations and soil Cu extractability tests are often poor as shoot concentrations will typically be unreflective of the extractability or solubility of Cu (Menzies et al., 2007). Therefore, root concentrations may provide stronger relationships with extractable Cu and this relationship has been suggested as an alternative to shoot concentrations (Chaignon et al., 2003). Other studies examining the success of predictions of Cu uptake by wheat using methods such as DGT and free ion activity
75 also found that these yielded poor relationships and that total Cu gave the best fit (Menzies et al., 2007; Nolan et al., 2005)
Responses for Cd and Zn were similar with Ca(NO3)2 and EDTA extractable concentrations providing the best relationships with shoot concentrations (r2 = 0.54 and r2 = 0.63, Figure 4.4 and Figure 4.7). These results indicate that shoot accumulation is moderately responsive to the availability of Cd and Zn as measured by Ca(NO3)2 and EDTA. However, from a predictive perspective, correlation results for Cd and Zn methods need improving and this may be achieved by either including soluble inorganic and/or organic complexes of these metals, or sorption suppression by competing cations (Antonladis and Tsadilas, 2007; Degryse et al., 2009b). The overall success of Ca(NO3)2 to predict shoot accumulation of Cd, Ni and Zn does
support the theory that weak salt extractants, such as Ca(NO3)2 are based on the assumption that bioavailable trace elements are located on the mineral surface and can be easily displaced by other desorbing cations such as Ca (Andrews et al., 1996; McLaughlin et al., 2006).
Bioavailability estimated by CE and free ion activity typically provided the poorest correlations with shoot concentrations (although results for Ni2+ activity were reasonable). This result may be partly explained by plant uptake of metal complexes and/or the dissociation of these complexes at the plant root-soil interface, which are not accurately accounted for by these methods. Furthermore, a study by Berkelaar and Hale (2003) found that plants either take up the complexes, or by alleviation of diffusional constraints for uptake of the free ion, through dissociation of labile metal complexes in the diffusive boundary. Also, in another experiment using the same
wheat variety the relationship between CE and free ion activity of Cd, Cu, Ni and Zn showed that the uptake of these metals was not only controlled by the free ion form, but it is also dependent on the concentrations of metal ligand complexes (Degryse et al., 2006a). These poor relationships could also be attributed to the large variations in measured concentrations that can arise when quantifying very small pools of metals in a complex and dynamic system, which would make obtaining a strong correlation difficult.
The success of these tests were not influenced by biosolids amendments which was evident in the examination of the relationship between Ca(NO3)2, the CE of metals and
76 shoot concentrations (Figure 4.15 and 4.16). This result concurs with some recent studies (McBride et al., 2009; Menzies et al., 2007), which provides evidence to the hypothesis that the bioavailable metals, especially Ni are located on mineral surfaces (Andrews et al., 1996; McLaughlin et al., 1999).
4.5 Conclusions
Treating and incubating three contrasting soil types with metal salts in the presence and absence of biosolids provided a unique opportunity to observe the overarching effects of biosolids on soils and plants and the interaction between the two, with respect to metal chemistry. The addition of biosolids did not significantly alter the
availability of Cd, Cu, Ni and Zn as determined by EDTA, Ca(NO3)2, soil solution, DGT and free ion activity. However, the presence of biosolids did result in significant increases of salinity, DOC, total Cu and Zn, as well as concentrations of Ca and Mg.
Dry matter production was adversely affected by increased levels of salinity, which inturn was significantly influenced by the presence of biosolids. Concentrations of Cu and Ni also negatively impacted on dry weight however these were not as strongly correlated as levels of salinity were with dry weight. Concentrations of Cd in shoots were significantly lower in biosolids amended soils, irrespective of soil incubation time.
From a predictive perspective, the best overall measure of Cd, Zn and especially Ni concentrations in wheat plants was Ca(NO3)2-extractable soil metal. Measures of soil
solution, CE and free ion activity were poorly correlated with Cd and Zn shoot concentrations. Of the four metals, Ni consistently yielded the strongest correlations with all methods, while predictions of Cu concentrations in shoots were typically poor. This suggests that while Ni concentrations in wheat shoots are highly reflective of Ni solubility, shoot concentrations of Cu are non-responsive to the availability of Cu. Examination of the relationships between metal shoot concentrations and measures of metal availability for biosolids amended and unamended metal spiked soils revealed that the addition of biosolids did not alter the success of these tests. As this study was carried out on a range of soils, with varying metal concentrations and
77 biosolids, these findings support the conclusion that Ca(NO3)2 extraction is a robust method to predict Cd and Ni bioavailability.
78 Chapter 5 Evaluating the Effects of Biosolids and Metal Salt Applications on the Bioavailability of Cd, Cu, Ni and Zn in Soils Managed Under Field Conditions.
5.1 Introduction
Field studies using intact soil monolith lysimeters have been used to study the effects of soil amendments on the plant uptake and mobility of nutrients and non essential trace elements (Antoniadis, 2008; Deram et al., 2006; Goss and Ehlers, 2009; McLaren et al., 2003; 2004; 2005; Ruttens et al., 2006). Previous work carried out by Lincoln University and Environmental Science and Research Ltd. have used lysimeters as well as field trials to study the effects of amending soils with biosolids and metal salts on bioavailability (McLaren et al., 2003; 2004; 2005; Speir et al., 2004). One main advantage of conducting field trials over environmentally controlled pot trials is that soil and pasture are studied under more natural conditions, thereby potentially increasing the applicability to soil use and management practices (Goss and Ehlers, 2009).
The same rationale for examining the efficacy of biosolids amendment on metal contaminated soils outlined in Chapter 4 was employed in this study. However, the study in Chapter 4 was carried out under controlled environmental conditions in pots, while the study described in this chapter has been modified to examine the effects of biosolids amendments, metal salt applications, and time on soil chemistry, metal availability and plant responses grown in free draining soils managed under field conditions.
The main objective of this study was to examine the effects of biosolids amendment and metal salt applications on the bioavailability of Cd, Cu, Ni and Zn in contrasting soil types managed under field conditions over a 24-month period. The data generated was also used to assess the validity of measures developed to predict shoot
79 concentrations of Cd, Cu, Ni and Zn in a diverse range of New Zealand soil types, which is presented in Chapter 6.
5.2 Methods
5.2.1 Set-up of lysimeter experiment and treatment of soils
Twenty one lysimeters of approximately 20 cm diameter and 30 cm depth were collected from each of the Halkett (HK), Summit (SM) and Wakanui (WK) (Table 3.1). The design of the lysimeter castings and the methods of soil sampling were undertaken according to methods referenced in Cameron et al. (1992) and McLaren et al. (2003; 2004; 2005). Sixty three lysimeters in total were transported back to Lincoln University and placed randomly around a dug-out trench, located at the Lincoln University field laboratory (Figure 5.1). Lysimeters were housed in-ground to provide thermal insulation and natural environmental exposure.
80 MM C A C LM Lysimeter key HM C Halkett soil HM B Summit soil MM BLM Cd B Wakanui soil HM LM C B Treatment abbreviations Cd MM C = control Cd BLM B = biosolids only C BLM BLM = biosolids and low metals B LM = low metals LM MM = medium metals BLM MM HM = high metals B HM Cd = Cd only C LM BLM MM B HM BLM HM LM Cd LM Cd MM C LM Cd HM C MM B Cd BLM HM C Cd LM B MM B HM BLM C LM C B Cd MM C BLM B
Figure 5.1 Map view and photo of field lysimeter set-up. Locations of soil and treatments within the trench are also shown.
5.2.2 Soil treatments
In October 2006, the top 10 cm of each soil monolith was removed and thoroughly mixed with a single application of combined increasing concentrations of Cd, Cu, Ni and Zn and one of Cd only (Table 5.1). As with the treatment applications in Chapter 4, metals were added as hydrated metal sulphates. Biosolids, as received, were
81 sourced from the Christchurch City Council treatment works at Bromley and added to untreated soils and also to the lowest metal treatment concentration only at a rate equivalent to 400 kg N ha-1 (fresh biosolids collected for this study comprised on average 5.6% of N). The rate of biosolids application was based on the maximum rate of application to agricultural soils as outlined in the guidelines regarding the safe application of biosolids to land (NZWWA, 2003). Cadmium and Ni concentrations chosen were also based on current biosolids guideline levels (NZWWA, 2003). Copper and Zn concentrations were chosen based on previous complimentary work by Environmental Science and Research Ltd. and Lincoln University (McLaren and Clucas, 2001; Speir et al., 2007). Ryegrass (Lolium perenne, sourced from Luisetti Seed Grain Merchant, Rangiora, New Zealand) was sown immediately after treatment application. During January to March, 40 mm of water was added five times per week to maintain plant growth.
Table 5.1 Soil treatments for field lysimeter experiment. Metals were added in the form of sulphates and biosolids were applied at a rate equivalent to 400 kg N ha-1
Treatment Abbreviations Cd added Cu added Ni added Zn added (mg kg-1) (mg kg-1) (mg kg-1) (mg kg-1)
Control C - - - -
Biosolids B - - - - Biosolids + BLM 1 30 50 70 low metals Low metals LM 1 30 50 70 Medium MM 5 200 150 300 metals High metals HM 10 750 300 1000
Cd only Cd 10 - - -
82
5.2.3 Harvesting of ryegrass and sampling of soil
Ryegrass was harvested six months after treatment applications and thereafter at six- monthly intervals. A mixture of commerical grade superphosphate at a rate equivalent to 0.078 kg m-2 and urea at a rate equivalent to 0.042 kg m-2 was applied to the soil monoliths after each harvest. Immediately after harvest, the top 10 cm of the soil was sub-sampled (one core per lysimeter) using a steel corer, approximately 2.5 cm in diameter and 15 cm long. Sampled soil was air dried at 25ºC and then passed through a 2 mm sieve prior to analysis.
5.2.4 Analysis of plant and soil samples
Fresh plant cuttings were weighed to record the fresh weights. Plant cuttings were dried at 45°C removed and re-weighed. Dried plants were ground into small pieces using an A10 Yellowline plant grinder (IKA Ltd. Germany) then digested using 10 ml of concentrated HNO3 and placed on a heating block cycle (section 3.3.1). Digested plant samples were then analysed for total Cd, Cu, Ni and Zn using ICP-OES (section 3.3.1).
Soils were also analysed for total, EDTA extractable, and Ca(NO3)2 extractable metal concentrations (section 3.3.2.2 and 3.3.2.3). Extraction of soil solution and diffusive gradient in thin films (DGT) placement (section 3.3.2.4 and 3.3.2.5) were undertaken simultaneously on the remaining soil which had been placed in plastic pots and wetted to 90 ± 10 % of the soils maximum water holding capacity and left to equilibrate for 24 hours (section 3.2). Soil pH and soil solution pH were also measured (section 3.2). Concentrations of DOC, major anions and cations in the soil solution were also measured (section 3.3.2.4). Salinity of the soil solution was defined as the sum of Cl-, + 2- Na and SO4 concentrations (Taiz et al., 1998).
Speciation and free ion activity of metals in soil solution were determined using
WHAM 6.0. (Tipping, 1998). Effective concentration (CE) was calculated from metal
83 fluxes measured using DGT devices placed on soils and the supplementary modelling freeware 2D DIFS (Harper et al., 2000; Sochaczewski et al., 2007). A full description of the calculations and methods involved in DGT –DIFS and speciation modelling using WHAM 6.0. were given in sections 3.3.2.5 and 3.3.2.6. All soil extractions were analysed using ICP-OES (section 3.3.2).
5.2.5 Derivation of critical levels of metals in plant tissues
Critical levels (CL) of Cd, Cu, Ni and Z in ryegrass shoot tissue used to compare against plant results in this chapter were obtained from Macnicol and Beckett (1985). Critical levels are based on a statistically significant 10 % reduction in dry matter yield.
5.2.6 Statistical analyses
Comparisons of shoot metal concentrations, soil chemistry and assays of bioavailability between biosolids and non-biosolids amended soils used paired t-tests (Sigma Plot 8.02, SPSS, UK). Linear regressions were also carried out using Sigma Plot 8.02. Differences in correlation determination values between biosolids amended soils and unamended soils used grouped and separate linear regression analyses performed in GenStat 11 (VSN International, UK). The effect of time on metal availability and shoot metal concentrations used a generalised linear model performed in S-Plus 4.5 (Mathsoft, U.S.A.).
84 5.3 Results
5.3.1 Changes in soil chemistry
5.3.1.1 Dissolved organic carbon, pH and major anions and cations
Comparisons between biosolids amended (BLM) and unamended (LM) metal spiked soils revealed significant increases in concentrations of DOC, Ca and Mg in soil solution as well as an increase in salinity in soils amended with biosolids (Table 5.2). All soil solution variables shown in Table 5.2 changed significantly with time (Table 5.4). There were significant decreases in soil pH, soil solution pH and DOC, and of these, soil solution pH showed the highest overall rate of decrease (Table 5.3 and 5.4). In contrast, soil solution salinity and concentrations of Ca and Mg increased with time, of which Ca in soil solution showed the highest rate of increase (Table 5.3 and 5.4)
Table 5.2 Significance levels for comparisons of general soil parameters between pooled results for biosolids-amended (BLM) and unamended (LM) metal spiked soils (n = 27).
Differences were considered significant at P<0.05 and highlighted in bold (*data log10- transformed).
Variable P-value Soil pH 0.564 Soli solution pH 0.091 DOC (mg L-1) 0.030 Salinity of soil solution (mg L-1) * 0.007 Soil solution Ca concentration (mg L-1) * 0.008 Soil solution Mg concentration (mg L-1) * 0.027
85 Table 5.3 Mean soil chemistry characteristics from pooled treated soils, sampled six months, 12 months and 24 months after treatment application (values are means, n = 63).
Variable 6 months 12 months 24 months Soil pH 5.59 5.71 5.19 Soli solution pH 5.36 5.28 4.53 DOC (mg L-1) 143.3 224.8 101.4 Salinity of soil solution (mg L-1) 62.6 75.7 127.3 Soil solution Ca concentration (mg L-1) 27.8 53.2 113.5 Soil solution Mg concentration (mg L-1) 4.54 9.54 16.58
Table 5.4 Effect of time calculated from residual deviance (variation due to treatment effects removed) on soil chemistry charactersitics. Effects were considered significant at P<0.05 and highlighted in bold (*data log10-transformed). P-value for the T distribution is also included.
Variable P-value for Residual Degrees of P-value Gradient T deviance freedom distribution Soil pH 34.85 181 <0.001 -0.200 0.841 Soli solution pH 64.72 181 <0.001 -0.418 0.679 DOC (mg L-1)* 74.58 181 0.013 -0.233 0.818 Salinity of soil solution 60.23 181 <0.001 0.362 0.717 (mg L-1)* Soil solution Ca 147.5 181 <0.001 0.6599 0.515 concentration (mg L-1)* Soil solution Mg 131.9 181 <0.001 0.609 0.543 concentration (mg L-1)*
5.3.1.2 Metal extractability
Results from the pooled data showed that the presence of biosolids had a significant effect on the concentration of metal extracted using EDTA, CE and modelled free ion activity (Table 5.5). The addition of biosolids resulted in an increase in the EDTA- 2+ extractable Cd and Ni, CE-Zn, as well as modelled Cu activity (Table 5.5). In contrast the amount of CE-Cd and CE-Ni as measured by DGT decreased significantly in biosolids amended soils compared to unamended soils. Moreover, in the case of Ni,
86 the amount of Ni from the biosolids amended soils was less than half of the amount extracted from the unamended soils. The modelled activities of Ni2+ and Zn2+ also significantly decreased in the presence of biosolids (Table 5.5).
Table 5.5 Significance levels for comparisons of six potential measures of metal bioavailability between pooled sampling results for biosolids amended (BLM) and unamended (LM) metal spiked soils (log transformed data, n = 27). Differences were considered significant at P<0.05 and highlighted in bold.
Method Cd Cu Ni Zn Total metal 0.754 0.626 0.993 0.525 EDTA 0.013 0.363 0.001 0.147
Ca(NO3)2 0.362 0.088 0.498 0.464 Soil solution 0.272 0.657 0.087 0.093 DGT 0.030 0.143 0.025 0.037 Free ion activity 0.869 0.009 0.038 0.033
Cadmium: Pooled soil results revealed overall there was little significant change in the availability of Cd from soil with time. Of the six measures of Cd extractability, only soil solution Cd and Cd2+ activity significantly increased with time (Table 5.6 and 5.7), with Cd2+ activity having a higher rate of increase than concentrations of Cd in solution.
Copper: As with Cd, there were only two significant effects of time on on Cu 2+ availability in the soil. CE-Cu as measured by DGT and Cu activity increased significantly over time, with Cu2+ activity showing the highest rate of increase (Table 5.8 and 5.9).
Nickel: Of the six measures of Ni availability in soil, only Ni2+ activity increased significantly over time (Table 5.10 and 5.11).
Zinc: Compared with other metals tested Zn had the most significant changes in
metal extractability and solubility with time (Table 5.12 and 5.13). Total-Zn and CE- Zn both decreased significantly with time since treatment application, whilst soil solution-Zn and Zn2+ activity increased significantly with time. Rates of increase in
87 Zn2+ activity were similar to those of Cd2+ activity (Table 5.7) and Ni2+ activity (Table 5.11).
Table 5.6 Means for six potential measures of Cd bioavailability taken from pooled soil results from each sampling occasion (n = 63 for each sampling occasion).
Method 6 months 12 months 24 months Total Cd (mg kg-1) 14.2 12.8 15.2 EDTA (mg kg-1) 17.0 15.1 27.3 -1 Ca(NO3)2(mg kg ) 7.173 6.130 6.37 Soil solution (mg L-1) 1.33 0.17 0.43 DGT (mmol L-1) 1.01 0.76 1.40 Cd2+ activity (mol L-1) 0.44 x 10-6 0.62 x 10-6 0.15 x 10-5
Table 5.7 Effect of time as calculated from residual deviance (variation due to treatment effects removed) on six measures of Cd bioavailability. Effects were considered significant at
P<0.05 and highlighted in bold (*data log10-transformed). P-value for the T distribution is also included.
Method Residual Degrees of P-value for T- P-value Gradient deviance freedom Distribution Total Cd (mg kg-1) 52.46 181 0.427 0.029 0.976 EDTA (mg kg-1) 423.8 181 0.104 0.021 0.983 -1 Ca(NO3)2 (mg kg ) 107.3 181 0.1214 1.489 0.138 Soil solution (mg L-1) 132.4 181 0.002 0.344 0.731 DGT (mmol L-1) 430.7 181 0.758 -0.027 0.978 Cd2+ activity (mol L-1) 941.6 181 <0.001 0.784 0.434
88 Table 5.8 Means for six potential measures of Cu bioavailability taken from pooled soil results from each sampling occasion (n = 63 for each sampling occasion).
Method 6 months 12 months 24 months Total (mg kg-1) 481.0 410.0 474.0 EDTA (mg kg-1) 392.0 302.0 369.0 -1 Ca(NO3)2 (mg kg ) 46.10 31.10 48.70 Soil solution (mg kg-1) 3.51 1.72 5.03 DGT (mmol L-1) 29.8 9.3 26.1 Cu2+ activity (mol L-1) 1.90 x 10-5 2.56 x 10-6 3.09 x 10-5
Table 5.9 Effect of time as calculated from residual deviance (variation due to treatment effects removed) on six measures of Cu bioavailability. Effects were considered significant at
P<0.05 and highlighted in bold (*data log10-transformed). P-value for the T distribution is also included.
Method Residual Degrees of P-value for T- P-value Gradient deviance freedom Distribution Total (mg kg-1) 52.46 181 0.405 0.040 0.986 EDTA (mg kg-1) 423.8 181 0.239 0.161 0.872 -1 Ca(NO3)2 (mg kg ) 107.3 164 0.121 -0.119 0.905 Soil solution (mg kg-1) 132.4 181 0.152 0.110 0.912 DGT (mmol L-1) 430.7 181 <0.001 0.689 0.491 Cu2+ activity (mol L-1) 941.6 181 <0.001 1.436 0.150
Table 5.10 Means for six potential measures of Ni bioavailability taken from pooled soil results for each sampling occasion (n = 63 for each sampling occassion).
Method 6 months 12 months 24 months Total (mg kg-1) 70.3 65.4 54.4 EDTA (mg kg-1) 76.2 58.5 61.5 -1 Ca(NO3)2 (mg kg ) 46.6 39.8 31.2 Soil solution (mg L-1) 2.59 2.22 4.13 DGT (mmol L-1) 17.2 8.8 14.3 Ni2+ activity (mol L-1) 2.45 x 10-5 1.98 x 10-5 3.24 x 10-5
89 Table 5.11 Effect of time as calculated from residual deviance (variation due to treatment effects removed) on six measures of Ni bioavailability. Effects were considered significant at
P<0.05 and highlighted in bold (*data log10-transformed). P-value for the T distribution is also included.
Method Residual Degrees of P-value for T- P-value Gradient deviance freedom Distribution Total (mg kg-1) 30.68 181 0.208 -0.046 0.963 EDTA (mg kg-1) 378.8 181 0.228 0.156 0.876 -1 Ca(NO3)2 (mg kg ) 142.1 171 0.053 -0.168 0.866 Soil solution (mg L-1) 167.3 168 0.056 0.180 0.857 DGT (mmol L-1) 153 178 0.088 0.143 0.886 Ni2+ activity (mol L-1) 298.6 180 <0.001 0.742 0.459
Table 5.12 Means for six potential measures of Zn bioavailability taken from pooled soil results from each sampling occasion (n = 63 for each sampling occasion).
Method 6 months 12 months 24 months Total (mg kg-1) 209.6 184.0 167.3 EDTA (mg kg-1) 170.0 124.0 186.0 -1 Ca(NO3)2 (mg kg ) 95.0 124.0 186.0 Soil solution (mg L-1) 5.20 4.75 11.10 DGT (mmol L-1) 27.6 19.9 26.8 Zn2+ activity (mol L-1) 3.88 x 10-5 3.35 x 10-5 7.65 x 10-5
Table 5.13 Effect of time calculated from residual deviance (variation due to treatment effects removed) on six measures of Zn bioavailability. Effects were considered significant at P<0.05 and highlighted in bold (*data log10-transformed). P-value for the T distribution is also included.
Method Residual Degrees of P-value for T- P-value Gradient deviance freedom Distribution Total (mg kg-1) 23.63 181 0.039 -0.067 0.946 EDTA (mg kg-1) 225.8 181 0.009 0.261 0.764 -1 Ca(NO3)2 (mg kg ) 93.52 173 0.692 -0.027 0.878 Soil solution (mg L-1) 163 181 0.031 0.184 0.854 DGT (mmol L-1) 98.93 178 <0.001 -0.927 0.355 Zn2+ activity (mol L-1) 245.2 178 <0.001 0.774 0.439
90 Additionally, pooled soil solution compositions were taken 24 months after treatment application from 10 cm and 20cm depths. Comparisons betweent the depths found that with change in depth there was no significant difference in DOC levels (P = 0.434), nor pH (P= 0.235). However, there was a 10-fold reduction in soil solution Cd, Cu, Zn as well as Ca and Mg at depth (P <0.001), but no significant difference in Ni concentrations (P = 0.321). These results suggest that Ni in solution is more evenly distributed at least in the upper part (< 30 cm depth) of the soil horizon, while soluble Cd, Ca, Cu, Mg and Zn appear to largely be contained within the top 5 to 10 cm of the soil.
5.3.2 Responses of plants
Ryegrass growth in the treated lysimeters was well established by three months after treatment application in all but the two highest metal salt treatments (MM and HM). At the time of the six month harvest, ryegrass had yet to germinate in either MM or HM treatments across all soils (Figure 5.2). Subsequent harvestings of ryegrass from these treatments at 12, 18 and 24 months produced extremely low yields, and at 24 months there were still six lysimeters out of 63 in which ryegrass had still failed to germinate. In the MM and HM treated soils, eventual occurrence of grass growth in these lysimeters was restricted to the soil closest to the lysimeter edges. Ryegrass grown in the Cd-only treated lysimeters appeared to have shorter and thinner blades than those in the control, biosolids or low metal treated soils.
91 Cd-only treatment MM treatment
Figure 5.2 Apperance of ryegrass in two of the soil lysmeters six months after treatment application (MM = medium metal treatment).
5.3.2.1 Metal concentrations in ryegrass
A general comparison between the pooled data of ryegrass grown in biosolids amended and unamended metal treated soil over 24 months revealed that the addition of biosolids had no significant effect on the herbage concentrations of Cd (P = 0.054), Cu (P = 0.126), Ni (P = 0.076) or Zn (P = 0.167). Moreover, only plant concentrations of Ni (decreased) and Zn (increased) changed significantly with time (P < 0.001 for both Ni and Zn).
Cadmium: Concentrations of Cd generally increased with increasing treatment concentrations (Figure 5.3). However, the critical level (CL) for mature ryegrass of 30-35 mg kg-1 was not exceeded at any stage of the trial. Mean Cd concentrations in shoots over the experiment were not significantly affected by the presence of biosolids. However, a comparisons of pooled data at 24 months revealed a significant (P = 0.001) decrease in Cd concentrations in shoots of plants grown in biosolids
92 amended soils compared to unamended soils (Table 5.14). A comparison of Cd shoot concentrations over time revealed no significant changes (P = 0.188, Table 5.15).
Copper: There was little variation in Cu shoot concentrations with the CL for mature ryegrass of 30-35 mg kg-1 only exceeded at the highest metal treatment level (Figure 5.4). Although the presence of biosolids had no significant effect on the shoot concentrations of Cu over the 24 month trial, significant (P < 0.001) decreases in Cu shoot concentration were observed at 18 and 24 months in grass grown in biosolids amended soils (Table 5.14). A comparison of Cu shoot concentrations over the length of the trial revealed no significant change with time (P = 0.131, Table 5.15).
Nickel: The CL range of 130-220 mg kg-1 for mature ryegrass of Ni in shoots was exceeded in the two highest metal concentration treatments (Figure 5.5). The addition of biosolids had no significant effect on Ni shoot concentrations at any of the harvest times (Table 5.14). With the removal of variation due to treatment effects, results revealed that there was an overall significant (P < 0.001) decrease (0.65 mg kg-1 = antilog of -0.185, Table 5.15) in shoot Ni concentrations over the duration of the experiment.
Zinc: Generally shoot Zn concentrations increased with increasing treatment concentrations (Figure 5.6) and the CL range of 370-650 mg kg-1 for mature ryegrass was only exceeded in the highest metal treatment. Although the presence of biosolids had no significant effect on the overall concentration of Zn in shoots, there was a significant decrease observed in shoot Zn concentrations at the 24 month harvest in ryegrass grown in biosolids amended soils compared with unamended soils (P < 0.001, Table 5.14). Analysis of the shoot Zn concentrations over time revealed significant increase in shoot concentrations of Zn (P < 0.001, 1.32 mg kg-1 = antilog of 0.12, Table 5.15).
93 40 6 months 12 months 18 months CL 24 months ) -1
Key C = Control B = Biosolids 30 BLM = Biosolids + low metals 10 LM = Low metals MM = Medium metals HM = High metals Cd = Cd only Cd concentration(mg kg
0 CBBLMLMMMHMCd Treatment
Figure 5.3 Mean Cd concentrations in ryegrass shoot tissue (mg kg-1 dry weight) grown in biosolids amended and unamended metal spiked soils managed under field conditions over 24 months. The critical level (CL) for Cd as depicted by the red line is 30-35 mg kg-1 (Macnicol and Beckett, 1985). Bars denote the SEM, n = 45, 54, 56 and 59 for 6, 12, 18 and 24 months, respectively.
94 250 6 months 12 months 18 months 200 24 months ) -1 Key 150 C = Control B = Biosolids BLM = Biosolids + low metals LM = Low metals 100 MM = Medium metals HM = High metals Cd = Cd only Cu concentration (mg kg 50 CL
0 CBBLMLMMMHMCd Treatment
Figure 5.4 Mean Cu concentrations in ryegrass shoot tissue (mg kg-1 dry weight) grown in biosolids amended and unamended metal spiked soils managed under field conditions over 24 months. The critical level (CL) for Cu as depicted by the red line is 30-35 mg kg-1 (Macnicol and Beckett, 1985). Bars denote the SEM, n = 45, 54, 56 and 59 for 6, 12, 18 and 24 months, respectively.
95 300 6 months 12 months 250 18 months CL 24 months ) -1 200 Key C = Control B = Biosolids 150 BLM = Biosolids + low metals LM = Low metals MM = Medium metals 100 HM = High metals Cd = Cd only Niconcentration (mg kg
50
0 CBBLMLMMMHMCd Treatment
Figure 5.5 Mean Ni concentrations in ryegrass shoot tissue (mg kg-1 dry weight) grown in biosolids amended and unamended metal spiked soils managed under field conditions over 24 months. The critical level (CL) for Ni as depicted by the red line is 130-220 mg kg-1 (Macnicol and Beckett, 1985). Bars denote the SEM, n = 45, 54, 56 and 59 for 6, 12, 18 and 24 months, respectively.
96 600 CL 6 months 12 months 500 18 months 24 months ) -1
400 Key C = Control B = Biosolids 300 BLM = Biosolids + low metals LM = Low metals MM = Medium metals HM = High metals 200 Cd = Cd only Zn concentration (mg kg concentration Zn
100
0 CBBLMLMMMHMCd Treatment
Figure 5.6 Mean Zn concentrations in ryegrass shoot tissue (mg kg-1 dry weight) grown in biosolids amended and unamended metal spiked soils managed under field conditions over 24 months. The critical level (CL) for Zn as depicted by the red line is 370-560 mg kg-1 (Macnicol and Beckett, 1985). Bars denote the SEM, n = 45, 54, 56 and 59 for 6, 12, 18 and 24 months, respectively.
Table 5.14 Significance levels for comparisons between shoot metal concentrations in plants grown in biosolids-amended (BLM) and unamended (LM) metal spiked soils harvested after treatment application (log transformed data, n = 45, 54, 56 and 59 for 6, 12, 18 and 24 months, respectively). Differences were considered significant at P<0.05 and highlighted in bold.
Comparisons between biosolids amended and unamended metal spiked soils Metal Six months 12 months 18 months 24 months Cd 0.328 0.571 0.057 0.001 Cu 0.809 0.880 <0.001 <0.001 Ni 0.246 0.588 0.051 0.149 Zn 0.387 0.671 0.328 <0.001
97
Table 5.15 Effect of time calculated from residual deviance (variation due to treatment effects removed) on plant uptake of metals. Effects were considered significant at P<0.05 and highlighted in bold (*data log10-transformed). P-value for the T distribution is also included.
Metal Residual Degrees of P-value Gradient P-value for T- deviance freedom Distribution Cd 135.2 206 0.188 -0.069 0.968 Cu 108.9 214 0.131 -0.041 0.967 Ni 84.12 207 <0.001 -0.185 0.853 Zn 39.33 206 <0.001 0.120 0.904
5.3.2.2 Estimations of metal availability
Log-log relationships determined between pooled shoot metal concentrations and the six measures of metal availability taken from pooled soil results were significant (P < 0.001) for all four metals.
Cadmium: Total-Cd and soil solution Cd gave the strongest correlations with shoot Cd concentration of the six methods assessed (r2 = 0.73 and 0.71 respectively, Figure 5.7A and
D). The poorest correlation obtained was for EDTA-extractable and Ca(NO3)2-extractable Cd (r2 = 0.28 and 0.34 respectively Figure 5.7B and C) and shoot Cd concentration. Effective concentrations of Cd and Cd2+ activity gave very similar correlations with Cd shoot concentrations (r2 = 0.65 and 0.64, respectively, Figure 5.7E and F).
Copper: Correlations for Cu were the poorest of all metals (Figure 5.8). The highest r2 value obtained between Cu shoot concentrations and assay of Cu bioavailability was for Ca(NO3)2- extractable Cu (r2 = 0.59, Figure 5.8C), followed by total-Cu and soil solution-Cu (r2 = 0.49 for both measures, Figure 5.8and B). Of the six methods, Cu2+ activity gave the poorest correlation with shoot concentrations (r2 = 0.07, Figure 5.8F).
Nickel: Correlations for extractable and soluble Ni and shoot Ni concentrations were the strongest of the four metals (Figure 5.9). Of these, Ca(NO3)2-extractable Ni gave the strongest 2 2 correlation (r = 0.87, Figure 5.9C) closely followed by CE-Ni, soil solution Ni and total Ni (r = 0.84, 0.83 and 0.82 respectively, Figure 5.9E, D and A). EDTA-extractable Ni yielded the poorest correlation result with r2 = 0.44 (Figure 5.9B).
98 Zinc: Effective concentration of Zn gave the strongest correlation with shoot Zn concentrations (r2 = 0.76, Figure 5.10E), which was closely followed by Zn2+ activity, soil 2 solution Zn and Ca(NO3)2-extractable Zn (r = 0.68, 0.68 and 0.63 respectively, Figure 5.10F, D and C). As with Ni and Cd, EDTA extractable Zn gave the poorest correlation result with shoot Zn concentrations (r2 = 0.35, Figure 5.10B).
99 100 100 (A) (B)
10 10
1 1
0.1 0.1
0.01 0.01
2 2 y = -0.90 + 0.81x, r = 0.73, p> 0.001 y = 0.56 + 0.35x, r = 0.28, p<0.001 0.001 0.001 0.1 1 10 100 1000 0.001 0.01 0.1 1 10 100 1000
-1 -1 Log10 (Total-Cd, mg kg ) Log 10 (EDTA-Cd, mg kg )
100 100 (C) (D)
10 10
1 1 )
-1 0.1 0.1
0.01 0.01
2 2 y = -0.53 + 0.34x, r = 0.34, p<0.001 y = 0.46 + 0.53x, r = 0.71, p<0.001 0.001 0.001 0.01 0.1 1 10 100 1000 10-5 10-4 10-3 10-2 10-1 100 101 102
-1 -1 ( shoot ( shoot concentration mg kg
Log ) 10(Ca(NO3)2-Cd mg kg Log10 (Soil solution-Cd mg L ) 10 100 100 Log (E) (F)
10 10
1 1
0.1 0.1
0.01 0.01
2 2 y = 0.11 + 0.68x, r = 0.65, p<0.001 y = 2.69 + 0.41x, r = 0.64, p<0.001 0.001 0.001 0.0001 0.001 0.01 0.1 1 10 100 10-11 10-10 10-9 10-8 10-7 10-6 10-5 10-4 10-3
-1 Log 2+ activity mol L-1) Log10 (CE-Cd mmol L ) 10 (Cd
Figure 5.7 Relationships determined between soil extractable Cd and shoot Cd concentration using various extractants; (A) Total Cd, (B) EDTA-extractable Cd (C) Ca(NO3)2 extractable Cd, (D) soil solution, (E) effective concentration, (F) free ion activity. A linear regression line is fitted for each plot. n = 157.
100 100 100 (A) (B)
10 10
2 2 y = 0.48 + 0.23x, r = 0.49, p<0.001 y = 0.77 + 0.10x, r = 0.17, p< 0.001 1 1 1 10 100 1000 10000 0.01 0.1 1 10 100 1000 10000
-1 -1 Log10 (total-Cu, mg kg ) Log10 (EDTA-Cu, mg kg )
100 100 (C) (D) ) -1
10 10 (shoot concentration mg kg 10 2 2 Log y = 0.91 + 0.19x, r = 0.59, p<0.001 y = 1.07 + 0.19x, r = 0.49, p<0.001 1 1 0.01 0.1 1 10 100 1000 0.001 0.01 0.1 1 10 100
Log (Ca(NO ) -Cu, mg kg-1) -1 10 3 2 Log10 (Soil solution-Cu, mg kg )
100 100 (E) (F)
10 10
2 2 y = 0.91 + 0.12x, r = 0.25, p<0.001 y = 1.57 + 0.07x, r = 0.07, p<0.001 1 1 0.001 0.01 0.1 1 10 100 1000 10-14 10-13 10-12 10-11 10-10 10-9 10-8 10-7 10-6 10-5 10-4 10-3
-1 2+ -1 Log10 (CE-Cu, mmol L ) Log10 (Cu activity, mol L )
Figure 5.8 Relationships determined between soil extractable Cu and shoot Cu concentration using various methods; (A) Total Cu, (B) EDTA-extractable Cu (C) Ca(NO3)2 extractable Cu, (D) soil solution, (E) effective concentration, (F) free ion activity. A linear regression line is fitted for each plot. n = 157.
101 1000 1000 (A) (B)
100 100
10 10
1 1
2 2 y = -1.07 + 1.42x, r = 0.82, p< 0.001 y = 0.47 + 0.51x, r = 0.44, p< 0.001 0.1 0.1 1 10 100 1000 0.01 0.1 1 10 100 1000
-1 -1 Log10 (Total-Ni, mg kg ) Log10 (EDTA-Ni, mg kg )
1000 1000 (C) (D) ) -1
100 100
10 10
1 1 (shoot concentration, mg kg 10
Log 2 2 y = 0.71 + 0.58x, r = 0.87, p< 0.001 y = 1.52 + 0.60x, r = 0.83, p< 0.001 0.1 0.1 0.01 0.1 1 10 100 1000 0.0001 0.001 0.01 0.1 1 10 100
-1 -1 Log10 (Ca(NO3)2-Ni, mg kg ) Log10 (Soil solution-Ni, mg L )
1000 1000 (E) (F)
100 100
10 10
1 1
2 2 y = 1.04 + 0.65x, r = 0.84, p< 0.001 y = 4.00 + 0.48x, r = 0.75, p< 0.001 0.1 0.1 0.001 0.01 0.1 1 10 100 10-10 10-9 10-8 10-7 10-6 10-5 10-4 10-3
Log (C Ni, mmol L-1) 2+ -1 10 E Log10 (Ni activity, mol L )
Figure 5.9 Relationships determined between soil extractable Ni and shoot Ni concentration using various extractants; (A) Total Ni, (B) EDTA-extractable Ni (C) Ca(NO3)2 extractable Ni, (D) soil solution, (E) effective concentration, (F) free ion activity. A linear regression line is fitted for each plot. n = 157.
102 1000 1000 (A) (B)
100 100
2 2 y = 3.22 + 0.84x, r = 0.55, p<0.001 y = 1.28 + 0.28x, r = 0.35, p<0.001 10 10 10 100 1000 0.1 1 10 100 1000
-1 Log (EDTA-Ni, mg kg-1) Log10 (Total-Zn, mg kg ) 10
1000 1000 (C) (D) ) -1
100 100 (shoot concentration mg kg mg concentration (shoot 10 2 2
Log y = 1.35 + 0.34x, r = 0.63, p<0.001 y = 1.82 + 0.36x, r = 0.68, p<0.001 10 10 0.1 1 10 100 1000 0.01 0.1 1 10 100 -1 -1 Log10 (Soil solution-Zn, mg L ) Log10 (Ca(NO3)2-Zn mg kg )
1000 1000 (E) (F)
100 100
2 2 y = 1.55 + 0.40x, r = 0.76, p<0.001 y = 3.33 + 0.27x, r = 0.68, p<0.001 10 10 0.01 0.1 1 10 100 1000 10-9 10-8 10-7 10-6 10-5 10-4 10-3
-1 Log (Zn2+ activity, mol L-1) Log10 (CE Zn, mmol L ) 10
Figure 5.10 Relationships determined between soil extractable Zn and shoot Zn concentration using various extractants; (A) Total Zn, (B) EDTA-extractable Zn (C) Ca(NO3)2 extractable Zn, (D) soil solution, (E) effective concentration, (F) free ion activity. A linear regression line is fitted for each plot. n = 157.
103 5.3.2.3 Effects of biosolids amendments on relationships between soil extraction methods and metal concentration in plants
Evaluations of the correlation strengths determined between six potential measures of bioavailability and shoot metal concentrations (Figure 5.7-5.10) revealed that soil solution concentrations of metals gave the best overall estimate of bioavailability.
However results for Ca(NO3)2 extractable metals, EDTA-extractable metals, effective solution concentrations and free ion activities of metals were also used to explore any potential impacts that the addition of biosolids may have on the relationships obtained from pooled BLM and LM treatments.
Cadmium: Only two of the five methods gave overall significant relationships with shoot Cd concentrations (Table 5.16), namely soil solution and Cd2+ activity (P = 0.018 for both methods). Further analyses of gradients for all five methods revealed that biosolids did not affect the relationship determined between the measures of bioavailable Cd and shoot Cd concentrations (Table 5.16).
Copper: None of the five methods assessed for results from treatments BLM and LM gave significant relationships between the estimates of bioavalable metal concentrations and Cu concentrations in shoots (Table 5.17). Moreover, gradient analyses of the five methods revealed the addition of biosolids had no significant effect on the relationships determined between bioavailable Cu and shoot Cu concentrations (Table 5.17).
Nickel: Relationships determined between the overall available Ni concentrations and shoot concentrations of Ni were not significant (Table 5.18). Analyses of the gradients revealed that the addition of biosolids significantly affected the relationships determined between both soil solution Ni and Ni2+ activity and concentrations of Ni in shoots (P = 0.024 and 0.006 for soil solution Ni and Ni2+ activity respectively). Also the presence of biosolids significantly (P = 0.047) decreased the plant uptake of Ni in soil solution (Table 5.18).
Zinc: As with Cu, none of the five methods for estimating bioavailability gave significant relationships with shoot concentrations of Zn (Table 5.19). Analyses of the gradients also revealed that as with Cd and Cu the presence of biosolids does not significantly impact upon the relationship determined between shoot metal
104 concentrations and measures of bioavailability. Hence these results indicate that bioavailability of Cd, Cu and Zn as measured by these methods are not dependent upon substrate type.
Table 5.16 Significance levels obtained from grouped regression analyses for relationships determined between Cd concentrations in ryegrass and five potential measures of Cd availability in biosolids-amended (BLM) and unamended (LM) metal spiked soils. Differences were considered significant at P<0.05 and highlighted in bold (n = 27).
Method Overall P value Presence of biosolids Gradient difference (P-value) (P-value) EDTA 0.063 0.194 0.187
Ca(NO3)2 0.437 0.117 0.799 Soil solution 0.018 (r2 = 0.00) 0.376 0.293 DGT 0.312 0.227 0.910 Cd2+ activity 0.018 (r2 = 0.00) 0.317 0.860
Table 5.17 Significance levels obtained from grouped regression analyses for relationships determined between Cu concentrations in ryegrass and five potential measures of Cu availability in biosolids-amended (BLM) and unamended (LM) metal spiked soils. Differences were considered significant at P<0.05 and highlighted in bold (n = 27).
Method Overall P value Presence of biosolids Gradient difference (P-value) (P-value) EDTA 0.149 0.919 0.940
Ca(NO3)2 0.769 0.822 0.293 Soil solution 0.491 0.783 0.918 DGT 0.233 0.640 0.739 Cu2+ activity 0.134 0.659 0.733
105 Table 5.18 Significance levels obtained from grouped regression analyses for relationships determined between Ni concentrations in ryegrass and five potential measures of Ni availability in biosolids-amended (BLM) and unamended (LM) metal spiked soils. Differences were considered significant at P<0.05 and highlighted in bold (n = 27).
Method Overall P value Presence of biosolids Gradient difference (P-value) (P-value) EDTA 0.243 0.149 0.270
Ca(NO3)2 0.791 0.135 0.822 Soil solution 0.110 0.047 0.024 DGT 0.448 0.185 0.173 Ni2+ activity 0.171 0.057 0.006
Table 5.19 Significance levels obtained from grouped regression analyses for relationships determined between Zn concentrations in ryegrass and five potential measures of Zn availability in biosolids-amended (BLM) and unamended (LM) metal spiked soils. Differences were considered significant at P<0.05 and highlighted in bold (n = 27).
Method Overall P value Presence of biosolids Gradient differences (P- value) (P-value) EDTA 0.164 0.547 0.559
Ca(NO3)2 0.323 0.324 0.534 Soil solution 0.369 0.285 0.849 DGT 0.322 0.343 0.421 Zn2+ activity 0.296 0.278 0.573
5.4 Discussion
5.4.1 Influence of biosolids and time after treatment application on metal availability in soils
The addition of biosolids significantly increased the concentrations of DOC, salinity, Ca and Mg in soil solution as well as assays of bioavailability of all four metals. These results for increased DOC, salinity, Ca and Mg in soil solution concur with findings in Chapter 4. The length of time following treatment application had the most
106 effect on soil chemistry (Table 5.3). Over the duration of the field trial the pH of the soil and soil solution significantly decreased with time as did DOC (Table 5.2 and 5.4). Similar trends of decreasing pH and DOC with time have also been found in longer term field trials involving biosolids applications (Walter et al., 2002).
The increase of DOC in biosolids amended soils compared with unamended soils is most likley attributed to the degradation of biosolids (Table 5.2). Although the decomposition of biosolids provides extra DOC, leaching and mineralisation can lead to the reduction of DOC in soil solution with time (Amery et al., 2008; Antonidas and Alloway, 2002b), which may largely explain the results observed in this study.
The decomposition of organic matter has generally been accepted as a major cause of decreasing pH in soils amended with organic-rich composts (Helyar and Porter, 1989), However, in more recent studies describing the effects of organic soil amendments on soil chemistry Butterly et al. (2008) and Nwachaukwu and Pulford (2009) demonstrated that pH changes in soil amended with organic-rich material appear largely dependent on the type of organic material used. However, as the pH of the soil or soil solution was not significantly affected in the presence of biosolids (Table 5.2) a more likley explanation is that a combination of three of the following lead to decreasing soil pH overtime:
• decomposition of soil organic matter,
• plant activities (release of H+ as a response to the uptake of other nutrient cations);
• changes in the cation exchange capacity status of the soils.
Biosolids contain significantly (P<0.001) higher amounts of Ca, Mg and salinity levels than the soils resulting in higher levels of these in the biosolids amended soils compared with the unamended soils. While pH and DOC decreased with time, Ca, Mg and salinity of the soil solution increased with time (Table 5.2 and 5.4). Thus biosolids and length of time since treatment application resulted in increased amounts of Ca, Mg and salinity in the soil solution. Also six-monthly applications of superphosphate fertiliser after each harvest would have contributed to overall higher Ca levels. Increases in exchangeable Ca and in particular Mg with time can also be
107 attributed to the displacement of these cations (via competition from other cations) from exchange sites within the soil.
Biosolids amendment of metal spiked soils resulted in an increase in Cu2+ activity and also a decrease in the activities of Ni2+ and Zn2+ compared with unamended soils (Table 5.5). However, over the 24 month course of the field trial all four of the metal activities increased (Table 5.7, 5.9. 5.11 and 5.13). Along with increasing in free ion activities of each metal, the amount of Cd and Zn in soil solution also increased. This increase in Cd and Zn in soil solution could also be a results of the lowering of pH over time and this relationship between Cd and Zn solubility and soil pH has been demonstrated by many studies (Degryse et al., 2007; Gray and McLaren, 2006; Gray et al., 1999b; McBride et al., 1997; Nwachukwu and Pulford, 2009; Smith, 1994). The decrease in pH result in less available binding sites on soil colloids and clay materials for cations such as Zn and Cd, therefore leading to an increase of these metals in the solution phase (McBride, 1994).
The increase in free ion activities also coincided with general increasing soil solution metal concentrations. However, leaching, organic complexation, precipitation, adsorption and plant uptake of cations and anions can decrease the ionic strength of solution which could result in an increase of activities (Antoniadis and Alloway, 2002a; 2003a).
Of the four metals Zn was the only metal that revealed a significant loss in total Zn
from the top 10cm of the soil profile with time and a decrease in CE-Zn was also measured (Table 5.13). The loss of total Zn from the soil is most likely attributed to Zn leaching through the soil profile of the free-draining lysimeters. Losses of total soil Zn from the top of the soil profile has also been observed in a number of studies where biosolids were applied to soils managed under field conditions resulted in considerable movement of Zn through the soil profile (Antoniadis, 2008; Antoniadis and Alloway, 2002a; 2003a; McLaren et al., 2003; Speir et al., 2003b; Ruttens et al., 2006). Additionally, it was found that Zn was the most mobile of the metals examined out of Cd, Cu, Ni and Zn and these finding are overall in agreement with results described in this study. An increase in EDTA-extractable Zn with time was also observed and it may in part be attributed to the increased amount of Zn in more strongly complexed forms. This effect has also been described by Ashworth and
108 Alloway (2004) and Walter et al. (2002) as a consequence of increased soil contact time. This could also explain the increase in EDTA-extractable Cd and Ni observed in this study. Although, these type of complexation reactions are more pronounced at higher pH (Buekers et al., 2008).
Diffusive gradient in thin film method has alternatively been used to describe the degree of metal resupplied from the solid phase to the soil solution i.e the kinetics of metal resupply from solid phase to the soil solution (Zhang and Young, 2005; Zhang et al., 2004). Moreover, the rate of desorption measured by DGT, found that Cd and Zn rates are considerably faster than rates of Ni and Cu (Ernstberger et al., 2005) and this finding appeared similar to observations made in this study (Table 5.7, 5.11 and
5.13). Thus a decrease in the CE of Cd, Ni and Zn in soil amended with biosolids compared with unamended soils could be indicative of a decrease in the re-supply rate of these metals from the solid phase to the soil solution. This decrease in the resupply rate is most likely attributed to the presence of additional organic matter from biosolids which would increase the amount of binding sites available for cations (Kovarikova et al., 2007). In a study of Cd sorption in biosolids, Merrington and Smernik (2004) found that Cd sorption appeared to be dominated by the inorganic fraction, which may also contribute to the decrease in CE-Cd observed in this study.
5.4.2 Responses of plants grown in biosolids amended and unamended metal spiked soils
5.4.2.1 Presence of biosolids
Comparisons at the 18 and 24 month harvests yielded significant decreases in Cd, Cu and Zn shoot concentrations in soils amended with biosolids compared to unamended metal spiked soils (Table 5.14). The results for Cd agree with the significant changes
in CE-Cd (decrease, P = 0.030), indicative of reduced plant available Cd. Although there was an increase in EDTA extractable Cd (increase, P = 0.013) observed, this assay weakly correlated with shoot Cd concentration (r2 = 0.28) and thus EDTA- extractable Cd is a poor measure of plant available Cd.
109 Findings for shoot concentrations of Cu and Zn in plants grown in biosolids amended
soils compared with unamended soils (Table 5.14) are not reflected in th CE-Zn and Cu2+ activity increases observed in the presence of biosolids (Table 5.5). However no assay of Cu bioavailability was able to significantly predict greater than 50% of Cu concentrations in shoots, hence it is not surprising that shoot concentrations of Cu and
assays of Cu bioavailability are in disagreement. Although CE-Zn did give a good estimate of Zn concentrations in shoots (r2 = 0.76) using pooled data. It is likely differences between shoot concentrations of Zn in plants grown in biosolids amended soils compared with unamended soils at the 24 month harvest is more a result of time. Time since treatment application has been shown to significanty change bioavailable pools of Zn (Table 5.13).
In Chapter 4, DOC, salinity, exchangeable Ca and Mg were found in higher concentrations in biosolids amended soils compared with unamended soils, although no strong relationship was found between these variables and shoot concentrations of Cd. However, results for decrease shoot concentrations of Cd in the presence of biosolids did support the theory that biosolids have the capacity to retain Cd (Nwachukwu and Pulford, 2008; Ruttens et al., 2006; van Herwijnen et al., 2007a) This theory may partially account for the difference in shoot concentrations of Cd in plants grown in biosolids amended soils compared with unamended soils at the 24 month harvest ocassion (Table 5.14). However, the absence of such significant results from previous harvest occasions (Table 5.14) makes it more likely that changes in shoot concentrations of Cd and also Cu, Ni and Zn is most likely a result of time than the presence of biosolids.
5.4.2.2 Change in shoot metal concentrations over time
Shoot concentrations of Ni and Zn were the only metals that showed significant changes over time (Table 5.15). While concentrations of Zn in ryegrass increased on average by 1.32 mg kg-1 over 24 months Ni concentrations in shoots decreased by 0.65 mg kg-1. However, there was considerably more variation in the data for Ni than there was for Zn (Table 5.11 and 5.13). This result for Ni concentrations in shoots is in disagreement with extractable and soluble Ni results, where the only significant
110 change was in Ni2+ activity which increased with time (Table 5.11). Although
assessment of concentrations of Ca(NO3)2 extracted Ni and soil solution Ni did not yield significant results, p-values were only marginally higher than 0.05 (P = 0.053
and 0.056 for Ca(NO3)2 and soil solution Ni respectively). Significant results for soil solution Zn and Zn 2+ activity coincided with an increase in Zn uptake by ryegrass over 24 months (Table 5.13 and 5.15) but were in contrast to total Zn in the top 10 cm of the soil profile, which decreased significantly with time as did the CE of Zn (Table 5.13). These results suggest that the concentration of Zn in solution and free ion Zn is
more related to the bioavailable fraction than total Zn or CE-Zn. This result also coincides with a decrease in pH over time, which has been shown to significantly influence soil solution to plant transfer of Zn in ryegrass grown in metal contaminated soil (Hough et al., 2005)
Given that five of the soil assay of Ni bioavailability were able to predict greater than 75% of Ni concentrations in shoots (Figure 9), the discrepancy between assays of Ni bioavailability and plant uptake of Ni over time (Table 5.11 and 5.15) could be accounted for in two ways, namely cation competition and the species of Ni in solution. Species of Ni in solution (e.g. Ni complexed with soluble organic ligands), and competition from other cations (Meers et al., 2009) could limit Ni uptake (Chaney et al., 2008), even though soluble forms of Ni would be increasing with time. Analyses of the soil solution results certainly confirmed increases in Ca, Mg, Cu and Zn overtime (Table 5.3, 5.9 and 5.12). Although this study did not differentiate between soluble organic ligand types, in particular it has been demonstrated in other studies that Ni speciation in the presence of different types of dissolved organic carbon has a major impact on the availability of Ni (Antoniadis and Alloway, 2002b; Ashworth and Alloway, 2004; Doig and Liber, 2007; Molas and Baran, 2004; Nolan et al., 2009; Rooney et al., 2007b). In particular, it has been demonstrates that even when the solubility of Ni increases this does not necessarily translate into increased bioavailability or uptake by plants (Antoniadis and Alloway, 2002b; Rooney et al., 2007b).
111 5.4.3 Estimates of metal bioavailability
Findings for Ni in this study were in agreement with results from Chapter 4, indicating that shoot concentrations of Ni are highly indicative of the concentration of Ni present in solution and Ni weakly sorbed to soil particles. Hence, the solubility of
Ni as measured by Ca(NO3)2 extraction can reliably predict the plant availability and
phytotoxicity resulting from excess Ni in soil. The validity of Ca(NO3)2 extractable Ni is further demonstrated given that the robustness of the assay was determined despite confounding factors such as:
• use of contrasting soils spiked with metal salts in the presence and absence of biosolids;
• managed under controlled and field conditions;
• use of two different plant species.
The range of data in the literature regarding the success of a neutral salt extraction such as Ca(NO3)2 in measuring the bioavailable soil fraction is limited. Although
Ca(NO3)2 extraction has provided the most useful indication of metal bioavailability across of range of soils and metal concentrations (Meers et al., 2009; Menzies et al., 2007; Wang et al., 2009).
The relationships for Cd and Zn concentrations in shoots and soil solution, CE and free ion activities were similar to each other and these methods provided the best relationships (Figure 5.7 andFigure 5.10). The similarity in behaviour of these metals
was also observed in Chapter 4. However, correlations for Ca(NO3)2-extractable Cd and shoot Cd concentrations were higher in the wheat seedling experiment (r2 = 0.54), than with Cd concentration in ryegrass (r2 = 0.34).
Comparing the correlation results for Zn in this study, with those obtained in the wheat seedling experiment, revealed a stronger relationship with ryegrass 2+ concentration and Ca(NO3)2 extracted Zn, with CE-Zn, soil solution and Zn activity also able to significantly describe greater variation in ryegrass shoot concentrations than wheat grown under controlled environments. These results suggest that ryegrass uptake of Cd and Zn in a field managed system is more related to Cd and Zn in
112 solution. Moreover the similar coefficient of determination values obtained for Cd and Zn concentrations in shoots and free ion activity (r2 = 0.64 and r2 = 0.68 for Cd and Zn, respectively), suggests that free ion Cd and Zn the most available component
present in the soil solution. Stronger relationships between CE-Zn and shoot Zn concentrations in ryegrass (r2 = 0.76) than wheat (r2 = 0.43) suggests a greater demand for Zn with increased growth in the longer running field trial compared to the pot trial, as there would be a greater need to replenish soil solution Zn as it is taken up by the plants. Coefficient of determination values for CE-Cd were generally lower
than for CE-Zn, which suggest competitive sorption may also partly explain the CE results, and this has also been demonstrated by Antonladis and Tsadilas (2007) where Zn was more strongly bound than Cd and Ni, with the order of reduced sorption being Zn > Ni > Cd. However, internal detoxification of Cd by the plant may also account for the poorer relationship between Cd measures of bioavailability and Cd concentrations in shoots compared with Zn which is an essential micronutrient (Reid et al., 2003a).
Relationships between measured soil concentrations of Cu and plant concentrations of Cu were the poorest of all the metals assessed (Figure 5.8). Results obtained in the wheat seedling experiment (Chapter 4) were also in agreement with this finding. The 2 strongest relationship obtained for Cu was provided by Ca(NO3)2 extractable Cu (r = 0.59), while in the wheat experiment, total Cu provided the strongest relationship (r2 = 0.15). Overall, relationships determined between extracted Cu and plant concentrations were higher in results obtained from the field lysimeter trial. The lack of strong relationships obtained between extractable/soluble Cu and plant concentrations across soil types was noted by Menzies et al. (2007) and Nolan et al. (2005). Zhang et al. (2001) showed that stronger relationships can be obtained 2 between the CE of Cu and plant uptake (r = 0.95), although this result was obtained from a smaller dataset than this study. However a known hyperaccumulating plant species was used in this study and soils were only contaminated with Cu. Further comparisons between assays of Cu bioavailability and plant uptake are explored in Chapter 6.
Interestingly, when relationships were determined between shoot concentrations of Cd and corresponding assays of Cd bioavailability in biosolids amended soils (BLM) and unamended soil (LM), only soil solution and free ion activity for Cd revealed
113 significant correlations (Table 5.16). Additionally, the relationships determined between these assays and shoot concentrations of Cd were not significantly affected by the presence of biosolids (Table 5.16). Conversely, McLaughlin et al. (2006) compared the solubility of Cd present in biosolids matrix versus soil spiked by Cd salts and found that the solubility of Cd was found to be equal or greater in soils when present bound in biosolids matrix than as Cd in a hydrated salt form which disagrees with findings in this study.
The presence of biosolids did significantly change the relationship determined between Ni concentrations in shoots and corresponding bioavailability assays of Ni in solution and Ni2+ (Table 5.18). These findings suggest that in biosolids amended soils, Ni uptake by ryegrass is lower from these measured bioavailable fractions compared to plant uptake of Ni from unamended soils. This trend was not observed in the wheat in Chapter 4, where a larger range of Ni concentrations in soils were used to determine relationships between extraction assay and plant concentrations. However, the lack of significant correlations for relationships between extracted Ni and corresponding concentrations in shoots may also be due to the limited range of Ni concentrations present in the soil and shoots. Although the robustness of these bioavailability measures in this experiment, appear dependent on the presence or absence of biosolids it is difficult to ascertain the effect of biosolids amendment has on correlation strengths from data with a limited range of values.
The relationship between EDTA, Ca(NO3)2 and the CE of each metal and shoot metal concentrations were not significantly different in the presence of biosolids. Given the
overall success of Ca(NO3)2 as a predictor of shoot metal concentration in this experiment and the wheat seedling experiment, it appears that this method is the most robust for predicting metal bioavailability across contrasting soil properties, varying metal concentrations and presence of biosolids. This lends support to the hypothesis that bioavailable metals are located on mineral surfaces and these can be displaced by competing cations (Andrews et al., 1996; McLaughlin et al., 1999).
114 5.4.4 Conclusions
Applying a treatment regime of increasing concentrations of metal salts in the presence and absence of biosolids to undisturbed soil monolith lysimeters of three contrasting soil types provided two important research opportunities:
• to examine the combined effects of biosolids amendment and metal salt applications on the soil chemistry and bioavailability of Cd, Cu, Ni and Zn in contrasting soil types managed under field conditions over a 24- month period;
• to provide a field comparison to the wheat seedling experiment carried out under controlled environmental condition as described in Chapter 4.
The addition of biosolids significantly increased the amount of DOC, salinity, Ca and Mg in solution and these results concurred with findings in Chapter 4. Biosolids also altered the extractability and solubility of Cd, Cu, Ni and Zn, although this was assay dependent, with Ni yielding the greatest number of significant differences, and CE and free ion activity yielding the greatest number of significant differences of the six methods used. Time had a larger overall influence on soil chemistry and metal bioavailability, with decreases in pH and DOC, and increases in salinity, Ca and Mg in soil solution. The free ion activities of each of the metals also increased with time, as did soil solution Cd and Zn, and the CE of Cu. Results for Zn suggested that a significant amount of Zn leaches through the soil profile with time. These findings suggests that length of experiments may alter the strength of the relationships between measures of metal bioavailability and shoot metal concentrations and thus experiments carried out over shorter time periods may not be appropriate use as models of longer term contamination studies.
Overall the addition of biosolids did not result in any significant changes in metal uptake by plants and only shoot concentrations of Ni and Zn significantly changed with time. The decrease in Ni concentration over time is most likely attributed to competition from other cations (i.e. Ca, Mg, Na and Zn). While increased Zn uptake may be linked to increased plant nutrient requirements a decrease pH could also result
115 in increased bioavailable forms of Zn and also competitive cation effects resulting from relatively high concentrations of Zn present in the soils.
Relationships determined between metal concentrations in soil solution and shoot metal concentrations showed the overall highest coefficent of determinations. However, in the presence of biosolids the relationship between shoot Ni concentration and Ni in soil solution was significantly affected as was the relationship between shoot Ni concentrations and Ni2+ activity. From a predictive perspective, the best overall bioavailable measure of Cd, Cu, Zn and especially Ni concentrations determined from a range of soils containing a range of metals in the presence and
absence of biosolids was Ca(NO3)2 extractable soil metal. Additionally, relationships
determined between Ca(NO3)2 extractable soil metal and ryegrass concentrations in this study were stronger than those obtained in Chapter 4.
116 Chapter 6 Evaluating Methods for Estimating the Bioavailability of Metals Using a Range of New Zealand Soils Amended with Different Rates of Biosolids and Metal Salts
6.1 Introduction
Numerous studies and several reviews on concepts and methods to accurately assess metal bioavailability and toxicity in soils have attempted, with some success, to link assays of metal extractability and solubility to plant tissue concentrations (Andrews et al., 1996; Basar, 2009; Degryse et al., 2009a; Lehto et al., 2006c; Lofts et al., 2007; McBride et al., 2009; McLaughlin et al., 2000; Meers et al., 2008; Menzies et al., 2007). A few studies have met with some success but in general results from these predictions have been largely inconclusive (Hamon et al. 1997; Hooda 2007; Zhang and Young 2005; Zhang et al. 2001).
Studies investigating the efficacy of metal bioavailability predictions have yet to establish the impact of organic amendments on the validity of these tests, especially across a range of soils. Part of the problem stems from inconsistencies in experimentation and insufficient datasets (Menzies et al., 2007). Moreover, very few studies have attempted to resolve the additional effect that different plant species may have on the outcome of these bioavailability measures (Almas et al., 2006; Datta and Young, 2005; Gray and McLaren, 2005; Lehto et al., 2006a).
The concentration of metal in the solution phase is thought of as the most bioavailable metal fraction with the amount sorbed onto the solid phase controlling the amount in solution (Degryse et al., 2009a; McBride et al., 1997a). While the free ion concentration is usually considered to be the major determinant species of bioavailability (Degryse et al., 2009a; 2009b), there is increasing interest in how low affinity ligands may affect metal bioavailability and toxicity (Antunes et al., 2006; Arnold et al., 2007; Stacey et al., 2008).
117 Several tecniques used to estimate the bioavailability of metals are based on the hypothesis that free ion and isotopically exchangeable metal fractions (those weakly bound to soil particles) reasonably represent the phytoavailable metal pool (McLaughlin et al., 1999; Meers et al., 2009; Menzies et al., 2007; Nolan et al., 2005; Zhao et al., 2006). Methods developed to estimate the bioavailable metal fraction have to consider factors that influence the size of the bioavailable metal pool, e.g. soil type, pH, dissolved organic carbon (DOC) forms and concentration, type and concentration of inorganic ligands, presence of competing cations, rhizophere processes, plant physiology etc (Alloway, 1995b; Basar, 2009; Gregory, 2006; Jones et al., 2004; McBride, 1994; Sauve, 2003), in order for them to accurately predict bioavailability. Because of the confounding factors it is unlikely that any given method will accommodate all possible factors and scenarios. The validity and strength of metal bioavailability predictions in soils is dependent on the suitability of the method used and the response of the biological model system being studied.
The main objective of this study was to assess the overall strength of the relationships determined between metal concentrations in plants and various measures of Cd, Cu, Ni and Zn solubility for a range of New Zealand soils amended with biosolids and/or combinations of metal salts. This was achieved by combining the data from the studies described in Chapters 4 and 5 with published and unpublished data form Environmental Science and Research Ltd. (ESR) in collaboration with Lincoln University.
In Chapters 4 and 5, wheat seedlings (Triticum aestivum) and perennial ryegrass (Lolium perenne) were used as the model plant systems. Both plant species were grown in three soil types spiked with increasing levels of metal salts, to which biosolids had been added to half of the metal treated soils at a rate equivalent to 400 kg N ha-1 (Chapter 4), and to the lowest metal salt concentration treatment only (Chapter 5).
This data was combined with data from the following field trials:
1. A seven-year field lysimeter study located at Lincoln University, Porirua and Hamilton in which biosolids have been applied annually from 2002 – 2008, to five soil types (different to soil types used in Chapters 4 and 5) at rates of 0,
118 200 and 800 kg N ha-1 in wet, composted and pelletised forms. Ryegrass (Lolium perenne) was re-sown annually on these plots (McLaren, 2005);
2. A pasture (Lolium perenne) field trial established at Lincoln University where Cu, Ni and Zn spiked biosolids were applied in wet form at a rate 400 kg N ha- 1 in 1997 and 1998; in 2003 these plots were subdivided into unlimed and limed treatments (McLaren, 2007);
3. Three pasture (Lolium perenne) field trials that involved addition of increasing rates of Cu and Zn were established in 2006 and 2007 and are located at Lincoln University, Nelson, South Island, New Zealand, and Hamilton, North Island, New Zealand (unpublished data);
4. A forest (Pinus radiata) field trial established at Doyles Block, Canterbury, involving addition of increasing concentrations of Cu, Ni and Zn spiked biosolids at a rate equivalent to 400 kg N ha-1 in 1998 and 1999 (McLaren et al., 2007).
Metal bioavailability in the studies by ESR and Lincoln University were determined by relating shoot concentrations to six different measures of bioavailability in a similar manner to Chapters 4 and 5. Correlations strengths from relationships determined between pooled soil and plant results were compared to establish which of these tests were most successful at predicting shoot concentrations. Differences between plant species relationships and assays of metal availability were also assessed. Comparisons between coefficients of determinations obtained for smaller soil treatment groups are discussed.
119
6.2 Methods
The data used in this Chapter comes from a range of previously conducted but closely related trials. The results from these trials have been pooled along with results from Chapters 4 and 5 to gain information on the effects of broad ranges of soil types, plant species and presence of biosolids has on asaays of metal bioavailability.
6.2.1 Set up of ESR and Lincoln University trials
All plant and soil samples were sampled, prepared and analysed according to methods previously described in Chapters 3, 4 and 5. A brief description of the trials by ESR and Lincoln University are given below.
6.2.1.1 Field lysimeter study to assess the effects of annual biosolids application to land
A lysimeter trial was established in 2002, and housed at Lincoln University (Figure 6.1), ESR Porirua and Landcare Research, Hamilton in which biosolids as received in composted and pelletised forms, were applied to five soil types (Table 3.1) annually at increasing rates of 0, 200 and 800 kg N ha-1.
Lysimeters were set-up according to Cameron et al. (1992). Each lysimeter was 50 cm in diameter and 70 cm deep. The procedures involved with the treatment and sampling of the lysimeters were described in detail by McLaren (2005). Pasture ryegrass was sown every year after biosolids application and is harvested once herbage has attained a height of 10cm. Harvested plant material was dried, weighed and ground for analysis (section 3.3.1). Soils were sub-sampled annually prior to biosolids application and analysed for total, EDTA-extractable, Ca(NO3)2 extractable metals (section 3.3.2).
120 In 2007, a wheat seedling bioassay (section 4.2.2) was carried out on soils sampled in October 2006. Metal concentrations in soil solution, DGT measured fluxes of metals and modeled free metal ion activities were also determined along with total, EDTA-
extractable and Ca(NO3)2 extractable metals.
Figure 6.1. Lysimeter set-up at Lincoln University to assess the effects on leachate quality, soil chemistry and herbage uptake of metals from annual application of biosolids.
6.2.1.2 Pasture field trial to assess the effects of biosolids spiked with metal salts on soil properties and plant uptake of metals
A pasture field trial assessing the effects of metal spiked biosolids was established in 1997 and 1998 at Lincoln University. Full details of the incorporation of metals into the biosolids was given in McLaren and Clucas (2001). In 1997 and 1998, applications of metal spiked biosolids were applied, at a rate equivalent to 400 kg N ha-1 yr-1 to approximately 1 m2 plots. Pastures were sown following biosolids incorporation into the soil. Pastures were sampled twice per year for analysis. Soil samples were collected annually and analysed for total, EDTA-extractable and
Ca(NO3)2-extractable metals (section 3.3.2). In 2003 these plots were subdivided into
121 unlimed and limed treatments and in 2006 a wheat seedling bioassay experiment (section 4.2.2) was conducted on soil sampled from the field trial. Metal concentrations in soil solution, DGT measured fluxes of metals and modeled free metal ion activities were also determined along with total, EDTA-extractable and
Ca(NO3)2 extractable metals. Both the field trial and wheat bioassay data are used in the analyses for this chapter.
6.2.1.3 Pasture field trials to assess the effects of metal salt applications on soil chemistry and plant uptake
Three pasture field trials were established in 2006 and 2007 at Lincoln University, Nelson, South Island, New Zealand (Figure 6.2) and Hamilton, North Island, New Zealand. All localities are shown in Figure 3.1. Increasing amounts of Cu (5, 20, 50, 120, 300, 750 and 2000 mg kg-1) and Zn (10, 30, 70, 160, 400, 1000 and 3000 mg kg- 1) in the form of hydrated metal sulphates incorporated into the top 10 cm of the soil at the commencement of the trial. Each plot was approximately 1 m2 in size.
Soils from each site were sampled twice per year between 2006 and 2007, and a wheat seedling bioassay (section 4.2.2) was carried out on the soils sampled in addition to analyses of herbage pasture. Sampled soils were analysed for total, EDTA-
extractable, Ca(NO3)2-extractable concentrations in soil solution, DGT measured fluxes of metals and modelled metal free ion activities (section 3.3.2).
122
Figure 6.2 Photograph of biosolids spiked with metal salts field trial, Top photo: Lincoln University trial set up in 1997, Bottom photo: metal salt trial set-up in 2006, Nelson, South Island, New Zealand.
6.2.1.4 Addition of metal spiked biosolids to forest soils
A forest field trial was established at Doyles Block in Canterbury, which involved the addition of Cu (at 50, 100, 150 and 200 mg kg-1), Ni (at 15, 30, 45 and 60 mg kg-1) and Zn (at 100, 200, 300 and 400 mg kg-1spiked biosolids at a rate equivalent to 400 kg N ha-1 in 1998 and 1999. Details of the incorporation of the metal salts into the biosolids were described in full in McLaren and Clucas (2001). Samples of soils were taken from the site six years after treatment application, wetted to 80% ± 5% field moisture capacity and incubated for two years at a constant 25°C. A wheat seedling bioassay (Chapter 4) was carried out on the incubated soils. Seedlings were harvested on day 20 of the experiment, prepared and analysed for total metals (section 4.2.2).
Soils were analysed for total, EDTA-extractable, Ca(NO3)2-extractable metals, metal
123 concentrations in soil solution, DGT measured fluxes of metals and modeled metal free ion activities (section 3.3.2).
6.2.2 Statistical analyses
Linear regression analyses used Sigma Plot 8.02 (Sigma Plot 8.02, SPSS, UK). Correlations between variables and gradient differences between wheat and ryegrass were determined using grouped and separate linear regression analyses performed in GenStat 11 (VSN International, UK). Coefficient of determination values (r2) less than 0.50 were considered poor measures of fit between variables.
124 6.3 Results
6.3.1 Effects of metal spiking of biosolids and soils on estimates of metal bioavailability
The range of shoot data variation that could be significantly (P <0.05) explained by the six assays of bioavailability in smaller more specific datasets was assessed by subdividing the pooled plant and soil data into distinct but closely related groups. Groupings were based on the soil treatment of no amendments (control), and different combinations of biosolids and metal salts. Subdivided groups are listed below and names in brackets were used in the subsequent description of results:
• control soils (control)
• soils having received an application of biosolids only (biosolids only)
• soils treated with metal spiked biosolids (spiked biosolids)
• metal spiked soils amended with biosolids (biosolids and metals)
• soils spiked with metal salts only (metals only)
Cadmium: None of the methods tested could significantly explain more than 50% of
variation in the data. Overall, Ca(NO3)2-extractable Cd (40%) and CE-Cd (21%) described the greatest amount of Cd shoot variation across all groups. However this was only observed in two of the groups (Table 6.1). For the biosolids only group, just two of the methods tested showed significant correlations with shoot Cd concentration but accounted for less than 10% of the shoot Cd concentration (Table 6.1). None of the methods were able to significantly explain any variation of Cd shoot concentration in spiked biosolids. In biosolids and metals, and metals only soils, just half of the methods tested had significant correlations with shoot concentrations but could explain less than 20 % variation in shoot Cd concentrations (Table 6.1).
Copper: Aside from the metal spiked biosolids group, no assay was able to describe greater than 50 % of the variation in shoot Cu concentrations (Table 6.2). Just two methods significantly correlated with shoot Cu concentrations in the control and biosolids only group, but could account for less than 20% of the variation in shoot Cu concentrations (Table 6.2). While results obtained for biosolids and metals were
125 significant for four of the assays tested, these accounted for less than 10% of variatioon in Cu shoot concentration. Four of the six assays tested were able to explain greater than 50% of shoot Cu concentration variability in the metal spiked biosolids group, with EDTA extractable Cu having the highest r2 value (0.63, Table 6.2). Correlation results for the metals only group were significant for all assays tested and had the second highest explanatory power to describe variation in shoot concentration of the assays tested compared to the metal spiked biosolids group (Table 6.2).
Nickel: Coefficients of determinations increased with increasing concentrations of Ni added to the soil. No method could explain greater than 50% of shoot Ni concentration variation in the control and biosolids only groups (Table 6.3). In the spiked biosolids group, all assays tested significantly correlated with shoot Ni
concentration but only (Ca(NO3)2-extractable Ni accounted for more than 50 % of Ni shoot concentration variability. In the biosolids and metals, four of the six assays could describe greater than 50% of shoot Ni concentration variability, with 2 (Ca(NO3)2-extractable Ni having the highest r value (0.81). In the metals only group, all assays gave significant correlations and could account for more than 50% of shoot
Ni concentration, with (Ca(NO3)2-extractable Ni able to describe 81 % of the variation in Ni shoot concentration.
Zinc: The number of significant relationships and r2 values increased with increased Zn concentrations across soil groups (Table 6.4). Only two of the assays in the control and biosolids only groups significantly correlated with shoot concentrations but accounted for less than 50% of shoot Zn variability. For spiked biosolids, biosolids and metals, and metals only, all methods significantly correlated with shoot concentration (Table 6.4). Five of the six assasys tested could explain more than 50% of the shoot Zn concentration variability in spiked biosolids group, with EDTA extractable Zn describing 87% of variability in shoot Zn concentration. In biosolids and metals, none of the assays tested could account for greater than 50% of variation in shoot Zn concentration. In the metals only group all methods could describe more
than 50% of shoot Zn concentration, with total Zn, Ca(NO3)2 extractable Zn and CE- Zn accounting for 70% of shoot Zn concentration variability.
126
Table 6.1 Coefficient of determination (r2) for each soil amendment group from relationships determined between plant concentration and six potential measures of Cd bioavailability. The highest scoring significant r2 value (P<0.05) for each group is highlighted in red. All data was log10 transformed for analysis. * P-values were greater than 0.05.
Method Control Biosolids only Spiked Biosolids and Metals only (n = 61) (n = 94) biosolids metals (n = 96) (n = 148) (n = 23) Total 0.04* 0.00* 0.00* 0.03* 0.01* EDTA 0.19 (P <0.001) 0.07 (P = 0.006) 0.05* 0.00* 0.00*
Ca(NO3)2 0.40 (P <0.001) 0.01* 0.05* 0.00* 0.00* Soil solution 0.05* 0.00* 0.09* 0.08 (P =0.003) 0.08 (P<0.001) DGT 0.00* 0.00* 0.00* 0.21 (P <0.001) 0.15 (P<0.001) Free ion 0.00* 0.04 (P = 0.04) 0.10* 0.07 (P =0.004) 0.05 (P=0.003) activity
Table 6.2 Coefficient of determination (r2) for each soil amendment group from relationships determined between plant concentration and six potential measures of Cu bioavailability. The highest scoring significant r2 value (P<0.05) for each group is highlighted in red. All data was log10 transformed for analysis. * P-values were greater than 0.05.
Method Control Biosolids only Spiked Biosolids and Metals (n = 82) (n = 102) biosolids metals (n = 96) only (n = 106) (n = 237) Total 0.41 0.00* 0.07 (P = 0.003) 0.57 (P<0.001) 0.06 (P = 0.005) (P<0.001) EDTA 0.30 0.00* 0.00* 0.63 (P<0.001) 0.05 (P = 0.014) (P<0.001)
Ca(NO3)2 0.43 0.00* 0.00* 0.23 (P<0.001) 0.03* (P<0.001) Soil solution 0.32 0.00* 0.01 (P<0.001) 0.52 (P<0.001) 0.02* (P<0.001) DGT 0.35 0.00* 0.00* 0.55 (P<0.001) 0.03 (P = 0.035) (P<0.001) Free ion 0.36 0.03* 0.18 (P<0.001) 0.39 (P<0.001) 0.01* activity (P<0.001)
127 Table 6.3 Coefficient of determination (r2) for each soil amendment group from relationships determined between plant concentration and six potential measures of Ni bioavailability. The highest scoring significant r2 value (P<0.05) for each group is highlighted in red. All data was log10 transformed for analysis. * P-values were greater than 0.05.
Method Control (n = Biosolids Spiked Biosolids and Metals only 76) only (n = 91) biosolids metals (n = 141) (n = 54) (n = 91) Total 0.20 0.67 0.72 0.00* 0.00* (P<0.001) (P<0.001) (P<0.001) EDTA 0.65 0.50 0.00* 0.01* 0.57 (P<0001) (P<0.001) (P<0.001)
Ca(NO3)2 0.07 0.07 0.31 0.81 0.81 (P=0.009) (P=0.005) (P<0.001) (P<0.001) (P<0.001) Soil solution 0.38 0.61 0.74 0.00* 0.00* (P<0.001) (P<0.001) (P<0.001) DGT 0.30 0.35 0.57 0.00* 0.00* (P<0.001) (P<0.001) (P<0.001) Free ion 0.09 0.36 0.43 0.63 0.00* activity (P=0.003) (P<0.001) (P<0.001) (P<0.001)
128
Table 6.4 Coefficient of determination (r2) for each soil amendment group from relationships determined between plant concentration and six potential measures of Zn bioavailability. The highest scoring significant r2 value (P<0.05) for each group is highlighted in red. All data was log10 transformed for analysis. * P-values were greater than 0.05.
Method Control Biosolids Spiked Biosolids and Metals only (n = 84) only (n = 98) biosolids metals (n = 297) (n = 38) (n = 95) Total 0.16 0.21 0.47 0.24 0.70 (P<0.001) (P<0.001) (P<0.001) (P<0.001) (P<0.001) EDTA 0.12 0.87 0.33 0.62 0.02* (P<0.001) (P<0.001) (P<0.001) (P<0.001)
Ca(NO3)2 0.46 0.65 0.36 0.70 0.02* (P<0.001) (P<0.001) (P<0.001) (P<0.001) Soil solution 0.12 0.62 0.23 0.64 0.00* (P<0.001) (P<0.001) (P<0.001) (P<0.001) DGT 0.61 0.33 0.70 0.00* 0.00* (P<0.001) (P<0.001) (P<0.001) Free ion 0.16 0.64 0.25 0.60 0.00* activity (P<0.001) (P<0.001) (P<0.001) (P<0.001)
6.3.2 Effects of plant species on estimates of bioavailability
Generally, concentrations of Cd and Zn were significantly (P<0.001) lower in ryegrass than wheat shoots, while Ni concentrations in ryegrass shoots were significantly higher than those measured in wheat shoots (P<0.001).
Cadmium: Comparisons of the relationships between shoot concentrations and the six potential measures of bioavailability for wheat and ryegrass (Figure 6.3) revealed significant differences between the plant species for each method assessed (Table 6.5). Wheat uptake was significantly higher in bioavailable fractions measured by total Cd, EDTA extraction, and Ca(NO3)2 extraction (Figure 6.3A-C and Table 6.5).
129 In contrast, ryegrass uptake of Cd was higher is bioavailable fractions measured by 2+ solution Cd, CE and Cd (Figure 6.3D-F and Table 6.5).
Copper: Relationships determined between shoot Cu concentration and six potential measures of bioavailability were significantly different between the two plant species, with most assays able to explain a greater amount of variation in the ryegrass data than in the wheat data (Figure 6.4 and Table 6.5). The difference between gradients revealed that Cu uptake by ryegrass was significantly less to that of wheat uptake of Cu across five of the assays of Cu bioavailability (Table 6.5).
Nickel: Apart from soil solution (P = 0.273), gradients for relationships between shoot Ni concentration and assays of Ni bioavailability differed significantly between the two plant species (Figure 6.5, and Table 6.5). Measures of Ni bioavailabiliy were able to describe more variation in ryegrass, than in wheat (Figure 6.5).
Zinc: Only two of the six potential measures of bioavailability assessed revealed significant differences in the gradients obtained for the two plant species (Figure 6.6 and Table 6.5). Results showed that Zn uptake by ryegrass was less in bioavailable
fractions as measured by EDTA-extraction and Ca(NO3)2 extraction than Zn uptake by wheat seedlings (Table 6.5).
130 100 100 (A) (B)
10 10
1 1
0.1 0.1
0.01 0.01 y = 0.11 + 0.68x, r2 = 0.65, p <0.001 (wheat) y = 0.04 + 0.71x, r2 = 0.65, p <0.001 (wheat) y = -0.90 + 0.81x, r2 = 0.72, p <0.001 (ryegrass) y = -0.56 + 0.35x, r2 = 0.28, p <0.001 (ryegrass) 0.001 0.001 0.01 0.1 1 10 100 1000 0.001 0.01 0.1 1 10 100 1000
Log (Total-Cd, mg kg-1) -1 10 Log10 (EDTA-Cd, mg kg ) 100 100 (C) (D)
) 10 10 -1
1 1
0.1 0.1
(shoot concentration,kg mg 0.01 0.01
10 10 y = 0.33 + 0.56x, r2 = 0.65, p <0.001 (wheat) y = 0.24 + 0.16x, r2 = 0.12, p <0.001 (wheat)
Log y = -0.53 + 0.34x, r2 = 0.34, p <0.001 (ryegrass) y = 0.50 + 0.54x, r2 = 0.72, p <0.001 (ryegrass) 0.001 0.001 10-5 10-4 10-3 10-2 10-1 100 101 102 103 10-9 10-8 10-7 10-6 10-5 10-4 10-3 10-2 10-1 100 101 102
-1 -1 Log10 (Ca(NO3)2-Cd, mg kg ) Log10 (Soil solution-Cd, mg L )
100 100 (E) (F)
10 10
1 1
0.1 0.1
0.01 0.01 2 y = 0.50 + 0.53x, r2 = 0.53, p <0.001 (wheat) y = 1.10 + 0.16x, r = 0.10, p <0.001 (wheat) 2 y = 0.11 + 0.68x, r2 = 0.65, p <0.001 (ryegrass) y = 2.67 + 0.41x, r = 0.64, p <0.001 (ryegrass) 0.001 0.001 0.0001 0.001 0.01 0.1 1 10 100 10-13 10-12 10-11 10-10 10-9 10-8 10-7 10-6 10-5 10-4 10-3 -1 Log (Cd2+ activity, mol L-1) Log10 (CE-Cd, mmol L ) 10
20 day old wheat seedlings Fitted line (wheat) Ryegrass pasture Fitted line (ryegrass)
Figure 6.3 Relationships determined between 20 day old wheat seedlings, pasture ryegrass concentrations and Cd concentrations obtained using various methods; (A) Total Cd, (B) EDTA- extractable Cd, (C) Ca(NO3)2 extractable Cd, (D) soil solution, (E) effective concentration, (F) free ion activity (n = 272 for wheat and n = 188 for ryegrass).
131 10000 10000 (A) (B)
1000 1000
100 100
10 10
1 1
0.1 0.1 y = 0.28 + 0.37x, r2 = 0.34, p <0.001 (wheat) y = 0.56 + 0.23x, r2 = 0.32, p <0.001 (wheat) y = 0.51 + 0.23x, r2 = 0.44, p <0.001 (ryegrass) y = 0.77 + 0.10x, r2 = 0.18, p <0.001 (ryegrass) 0.01 0.01 1 10 100 1000 10000 0.01 0.1 1 10 100 1000 10000 -1 -1 Log10 (EDTA-Cu, mg kg ) Log10 (Total-Cu, mg kg )
10000 10000 (C) (D) ) -1 1000 1000
100 100
10 10
1 1 (shoot concentration, mg kg 10 0.1 0.1 y = 0.99 + 0.22x, r2 = 0.30, p <0.001 (wheat) 2
Log y = 0.56 + 0.27x, r = 0.33, p <0.001 (wheat) y = 0.89 + 0.16x, r2 = 0.39, p <0.001 (ryegrass) y = 1.07 + 0.18x, r2 = 0.44, p <0.001 (ryegrass) 0.01 0.01 0.001 0.01 0.1 1 10 100 1000 0.0001 0.001 0.01 0.1 1 10 100 Log (Ca(NO ) -Cu, mg kg-1) -1 10 3 2 Log10 (Soil solution-Cu, mg L )
10000 10000 (E) (F)
1000 1000
100 100
10 10
1 1
0.1 0.1 y = 0.96 + 0.22x, r2 = 0.28, p <0.001 (wheat) y = 1.77 + 0.09x, r2 = 0.30, p <0.001 (wheat) 2 y = 0.89 + 0.12x, r2 = 0.28, p <0.001 (ryegrass) y = 1.45 + 0.06x, r = 0.40, p <0.001 (ryegrass) 0.01 0.01 0.001 0.01 0.1 1 10 100 1000 10-1710-1610-1510-1410-1310-1210-1110-10 10-9 10-8 10-7 10-6 10-5 10-4 10-3 10-2
-1 2+ -1 Log10 (CE-Cu, mmol L ) Log10 (Cu activity, mol L )
20 day old wheat seedlings Fitted line (wheat) Ryegrass pasture Fitted line (ryegrass)
Figure 6.4 Relationships determined between 20 day old wheat seedlings, pasture ryegrass concentrations and Cu concentrations obtained using various methods; (A) Total Cu, (B) EDTA-
extractable Cu, (C) Ca(NO3)2 extractable Cu, (D) soil solution, (E) effective concentration, (F) free ion activity for 20 day old wheat seedlings and ryegrass( n = 439 for wheat and n = 277 for ryegrass).
132 10000 10000 (A) (B)
1000 1000
100 100
10 10
1 1
0.1 0.1 2 y = -1.65 + 1.59x, r2 = 0.60, p <0.001 (wheat) y = -0.01 + 0.77x, r = 0.60, p <0.001 (wheat) 2 y = -0.96 + 1.36x, r2 = 0.80, p <0.001 (ryegrass) y = 0.47 + 0.51x, r = 0.44, p <0.001 (ryegrass) 0.01 0.01 1 10 100 1000 0.01 0.1 1 10 100 1000 -1 -1 Log10 (EDTA-Ni, mg kg ) Log10 (Total-Ni, mg kg )
10000 10000 (C) (D) )
-1 1000 1000
100 100
10 10
1 1 (shoot concentration, mg kg 10 0.1 2 0.1 2 Log y = 0.32 + 0.67x, r = 0.71, p <0.001 (wheat) y = 0.96+ 0.61x, r = 0.65, p <0.001 (wheat) y = 0.72 + 0.57x, r2 = 0.85, p <0.001 (ryegrass) y = 1.45 + 0.57x, r2 = 0.76, p <0.001 (ryegrass) 0.01 0.01 0.0001 0.001 0.01 0.1 1 10 100 1000 0.0001 0.001 0.01 0.1 1 10 100
-1 Log (Soil solution-Ni, mg L-1) Log10 (Ca(NO3)2-Ni, mg kg ) 10
10000 10000 (E) (F) 1000 1000
100 100
10 10 1
1 0.1
0.1 2 y = 0.43 + 0.49x, r2 = 0.50, p <0.001 (wheat) 0.01 y = 3.67 + 0.54x, r = 0.67, p <0.001 (wheat) y = 1.02 + 0.62x, r2 = 0.79, p <0.001 (ryegrass) y = 3.77 + 0.44x, r2 = 0.70, p <0.001 (ryegrass) 0.01 0.001 0.001 0.01 0.1 1 10 100 1000 10000 10-12 10-11 10-10 10-9 10-8 10-7 10-6 10-5 10-4 10-3
-1 Log (Ni2+ activity, mol L-1) Log10 (CE-Ni, mmol L ) 10
20 day old wheat seedlings Fitted line (wheat) Ryegrass pasture Fitted line (ryegrass)
Figure 6.5 Relationships determined between 20 day old wheat seedlings, pasture ryegrass concentrations and Ni concentrations obtained using various methods; (A) Total Ni, (B) EDTA- extractable Ni (C) Ca(NO3)2 extractable Ni, (D) soil solution, (E) effective concentration, (F) free ion activity (n = 295 for wheat and n = 208 for ryegrass).
133 105 105 (A) (B)
104 104
103 103
102 102
101 101 2 y = -0.09 + 0.94x, r2 = 0.63, p <0.001 (wheat) y = 1.10 + 0.52x, r = 0.67, p <0.001 (wheat) 2 y = 3.42 + 0.29x, r2 = 0.64, p <0.001 (ryegrass) y = 1.21 + 0.35x, r = 0.42, p <0.001 (ryegrass) 100 100 1 10 100 1000 10000 0.1 1 10 100 1000 10000 -1 -1 Log10 (Total-Zn, mg kg ) Log10 (EDTA-Zn, mg kg ) 105 105 (C) (D) ) -1 104 104
103 103
102 102 (shoot concentration, mg kg concentration, (shoot 1 1 10 10 10 2 2 Log y = 1.40 + 0.47x, r = 0.70, p <0.001 (wheat) y = 1.95 + 0.38x, r = 0.54, p <0.001 (wheat) y = 1.30 + 0.39x, r2 = 0.65, p <0.001 (ryegrass) y = 1.82 + 0.39x, r2 = 0.68, p <0.001 (ryegrass) 100 100 0.001 0.01 0.1 1 10 100 1000 10000 0.001 0.01 0.1 1 10 100 1000 10000 -1 -1 Log10 (Soil solution-Zn, mg L ) Log10 (Ca(NO3)2-Zn, mg kg )
105 105 (E) (F)
104 104
103 103
102 102
101 101 2 y = 1.72 + 0.41x, r2 = 0.58, p <0.001 (wheat) y = 3.48+ 0.28x, r = 0.49, p <0.001 (wheat) 2 y = 1.54 + 0.42x, r2 = 0.78, p <0.001 (ryegrass) y = 3.43 + 0.29x, r = 0.64, p <0.001 (ryegrass) 100 100 0.001 0.01 0.1 1 10 100 1000 10000 10-11 10-10 10-9 10-8 10-7 10-6 10-5 10-4 10-3 10-2 Log (C -Zn, mmol L-1) 2+ -1 10 E Log10 (Zn activity, mol L )
20 day old wheat seedlings Fitted line (wheat) Pasture ryegrass Fitted line (ryegrass)
Figure 6.6 Relationships determined between 20 day old wheat seedlings, pasture ryegrass concentrations and Zn concentrations obtained using various methods; (A) Total Zn, (B) EDTA- extractable Zn (C) Ca(NO3)2 extractable Zn, (D) soil solution, (E) effective concentration, (F) free ion activity (n = 439 for wheat and n = 220 for ryegrass).
134 Table 6.5 Significance levels and differences between plant species as determined from comparisons of relationships (gradients) between metal concentration in shoots of 20 day old wheat seedlings and ryegrass, and six potential measures of metal bioavailability. All data was log10 transformed for analysis. Differences were considered significant at P<0.05 and highlighted in bold.
Cadmium Method P-value for gradient ryegrass deviation from wheat comparison (reference gradient) Total 0.002 -0.2126 EDTA <0.001 -0.3602
Ca(NO3)2 <0.001 -0.2211 Soil solution <0.001 0.3806 DGT 0.007 0.1478 Free ion activity <0.001 0.2509 Copper Method P-value for gradient ryegrass deviation from wheat comparison (reference gradient) Total 0.432 -0.0206 EDTA <0.001 -0.1630
Ca(NO3)2 0.042 -0.0557 Soil solution 0.006 -0.0799 DGT <0.001 -0.1061 Free ion activity 0.001 -0.0345 Nickel Method P-value for gradient ryegrass deviation from wheat comparison (reference gradient) Total 0.013 -0.2363 EDTA <0.001 -0.2595
Ca(NO3)2 0.005 -0.0916 Soil solution 0.273 -0.0405 DGT 0.002 0.1364 Free ion activity <0.001 -0.0971 Zinc Method P-value for gradient ryegrass deviation from wheat comparison (reference gradient) Total 0.889 0.0078 EDTA <0.001 -0.1807
Ca(NO3)2 0.003 -0.0826 Soil solution 0.829 0.0068 DGT 0.742 0.0099 Free ion activity 0.730 0.0090
135
6.3.3 Estimations of bioavailability using pooled plant species and soil data
Log-log relationships determined between pooled plant metal concentrations and the six measures of metal bioavailability for pooled soil results were significant for all four metals (P < 0.001).
Cadmium: No assay tested was able to describe greater than 50% of the variability in shoot Cd concentrations (Figure 6.7). Effective Cd concentrations and total Cd gave the strongest relationship with shoot Cd concentrations of the six methods tested (r2 = 0.47 and r2 = 0.42, respectively Figure 6.7E and A). Ethylene diamine tetra-acetic acid
disodium salt - extractable Cd and Ca(NO3)2-extractable Cd gave similar correlations to each other (r2 = 0.33 and r= 0.31, respectively Figure 6.7B and C), while soil solution Cd and Cd2+ gave the poorest correlations (r2 = 0.19 and r2 = 0.1, respectively Figure 6.7D and F)
Copper: Overall, relationships between Cu concentrations in shoots and measures of availability were the poorest of the four metals assessed in this Chapter with none able to explain greater than 50% of variability in shoot Cu concentration (Figure 6.8). Of the assays tested, total Cu gave the strongest correlation with Cu shoot concentration 2 (r = 0.34, Figure 6.8A). Results for EDTA-extractable, Ca(NO3)2-extractable, soil 2+ 2 solution, CE and Cu had very similar results to each other ( r = 0.26 – 0.30 Figure 6.8B-F).
Nickel: Of the four metals examined, Ni generally gave the strongest relationships between shoot Ni concentration and the six methods assessed (Figure 6.9). Calcium nitrate extractable Ni gave the strongest correlation with shoot Ni concentration (r2 = 0.74 Figure 6.9C). Total and soil solution Ni had similar correlation responses to each other (r2 = 0.67 and r2 = 0.64, respectively Figure 6.9A and D). While Ni2+ activity 2 2 and CE had the poorest relationship with shoot Ni concentration (r = 0.55 and r = 0.54, respectively Figure 6.9F and E).
Zinc: As with Ni, Ca(NO3)2 gave the highest correlation with Zn concentrations in 2 2 shoots (r = 0.60 Figure 6.10C), followed by CE (r = 0.60 Figure 6.10E). Correlation
136 for determinations between EDTA-extractable and soil solution Zn concentrations were similar (r2 = 0.57 and r2 = 0.56, respectively Figure 6.10B and D), while total and Zn2+ activity had the same r2 values as each other (r2 = 0.52 Figure 6.10A and F).
137
100 100 (A) (B)
10 10
1 1
0.1 0.1
0.01 0.01
2 y = -0.43 + 0.70x, r2 = 0.42, p <0.001 y = -0.23 + 0.43x, r = 0.33, p <0.001 0.001 0.001 0.01 0.1 1 10 100 1000 0.001 0.01 0.1 1 10 100 1000 -1 -1 Log10 (EDTA-Cd, mg kg ) Log10 (Total-Cd, mg kg )
100 100 (C) X Data (D) ) -1 10 10
1 1
0.1 0.1 (shoot concentration, mg kg (
10 0.01 0.01 Log
y = -0.10 + 0.31x, r2 = 0.30, p <0.001 y = 0.23 + 0.22x, r2 = 0.19, p <0.001 0.001 0.001 10-5 10-4 10-3 10-2 10-1 100 101 102 103 10-9 10-8 10-7 10-6 10-5 10-4 10-3 10-2 10-1 100 101 102 Log (Ca(NO ) -Cd, mg kg-1) -1 10 3 2 Log10 (Soil solution-Cd, mg L )
100 100 (E) (F)
10 10
1 1
0.1 0.1
0.01 0.01
2 y = 0.35 + 0.50x, r2 = 0.47, p <0.001 y = 1.38 + 0.20x, r = 0.16, p <0.001 0.001 0.001 0.0001 0.001 0.01 0.1 1 10 100 10-13 10-12 10-11 10-10 10-9 10-8 10-7 10-6 10-5 10-4 10-3
-1 Log (Cd2+ activity, mol L-1) Log10 (CE-Cd, mmol L ) 10
Figure 6.7 Relationships determined between shoot Cd concentration and soil Cd concentrations obtained using various methods; (A) Total Cd, (B) EDTA-extractable Cd, (C)
Ca(NO3)2 extractable Cd, (D) soil solution, (E) effective concentration, (F) free ion activity. A linear regression line is fitted for each plot. n = 399.
138 10000 10000 (A) (B) 1000 1000
100 100
10 10
1 1
0.1 0.1 2 y = 0.37 + 0.31x, r = 0.34, p <0.001 y = 0.63 + 0.21x, r2 = 0.27, p <0.001 0.01 0.01 1 10 100 1000 10000 0.01 0.1 1 10 100 1000 10000
-1 -1 Log10 (Total-Cu, mg kg ) Log10 (EDTA-Cu, mg kg )
10000 10000 (C) (D) ) -1 1000 1000
100 100
10 10
1 1 (shoot concentration, mg kg 10 0.1 0.1 Log y = 0.95 + 0.19x, r2 = 0.30, p <0.001 y = 1.15 + 0.22x, r2 = 0.29, p <0.001 0.01 0.01 0.001 0.01 0.1 1 10 100 1000 0.0001 0.001 0.01 0.1 1 10 100 -1 -1 Log10 (Ca(NO3)2-Cu, mg kg ) Log10 (Soil solution-Cu, mg L )
10000 10000 (E) (F)
1000 1000
100 100
10 10
1 1
0.1 0.1
y = 0.92 + 0.18x, r2 = 0.26, p <0.001 y = 1.65 + 0.08x, r2 = 0.30, p <0.001 0.01 0.01 0.001 0.01 0.1 1 10 100 1000 10-1710-1610-1510-1410-1310-1210-1110-10 10-9 10-8 10-7 10-6 10-5 10-4 10-3 10-2
-1 Log (Cu2+, mol L-1) Log10 (CE-Cu, mmol L ) 10
Figure 6.8 Relationships determined between shoot Cu concentration and soil Cu concentrations obtained using various methods; (A) Total Cu, (B) EDTA-extractable Cu, (C)
Ca(NO3)2 extractable Cu, (D) soil solution, (E) effective concentration, (F) free ion activity. A linear regression line is fitted for each plot. n = 592.
139 10000 10000 (A) (B)
1000 1000
100 100
10 10
1 1
0.1 0.1 y = -1.47 + 1.56x, r2 = 0.67, p <0.001 y = 0.13 + 0.69x, r2 = 0.56, p <0.001 0.01 0.01 11010010000.01 0.1 1 10 100 1000
-1 Log (EDTA-Ni, mg kg-1) Log10 (Total-Ni, mg kg ) 10 10000 10000 (C) (D) ) -1 1000 1000
100 100
10 10
1 1 (shoot concentration,mg kg 10 0.1 0.1 Log y = 0.45 + 0.65x, r2 = 0.74, p <0.001 y = 1.17 + 0.61x, r2 = 0.64, p <0.001 0.01 0.01 0.0001 0.001 0.01 0.1 1 10 100 1000 0.0001 0.001 0.01 0.1 1 10 100
-1 -1 Log10 (Ca(NO3)2-Ni, mg kg ) Log10 (Soil solution-Ni, mg L )
10000 10000 (E) (F) 1000 1000
100 100
10 10 1
1 0.1
0.1 0.01 y = 0.66 + 0.55x, r2 = 0.54, p <0.001 y = 3.57 + 0.48x, r2 = 0.55, p <0.001 0.01 0.001 0.001 0.01 0.1 1 10 100 1000 10000 10-12 10-11 10-10 10-9 10-8 10-7 10-6 10-5 10-4 10-3
2+ -1 -1 Log10 (Ni activity, mmol L ) Log10 (CE-Ni, mmol L )
Figure 6.9 Relationships determined between shoot Ni concentration and soil Ni concentrations obtained using various methods; (A) Total Ni, (B) EDTA-extractable Ni, (C)
Ca(NO3)2 extractable Ni, (D) soil solution, (E) effective concentration, (F) free ion activity. A linear regression line is fitted for each plot. n =425.
140 105 105 (A) (B)
104 104
103 103
102 102
101 101
2 y = 3.46 + 0.28x, r = 0.52, p <0.001 y = 1.13 + 0.46x, r2 = 0.57, p <0.001 100 100 1 10 100 1000 10000 0.1 1 10 100 1000 10000 -1 -1 Log10 (EDTA-Zn, mg kg ) Log10 (Total-Zn, mg kg )
105 105 (C) (D) ) -1 104 104
103 103
102 102 (shoot concentration, mg kg concentration, (shoot 1 1 10 10 10 Log y = 1.37 + 0.44x, r2 = 0.65, p <0.001 y = 1.91 + 0.41x, r2 = 0.56, p <0.001 100 100 0.001 0.01 0.1 1 10 100 1000 10000 0.001 0.01 0.1 1 10 100 1000 10000 -1 Log (Soil solution-Zn, mg L-1) Log10 (CA(NO3)2-Zn, mg kg ) 10
105 105 (E) (F)
104 104
103 103
102 102
101 101
2 y = 1.66 + 0.41x, r = 0.60, p <0.001 y = 3.46 + 0.28x, r2 = 0.52, p <0.001 100 100 0.001 0.01 0.1 1 10 100 1000 10000 10-11 10-10 10-9 10-8 10-7 10-6 10-5 10-4 10-3 10-2 Log (C -Zn, mmol L-1) 2+ -1 10 E Log10 (Zn activity, mol L )
Figure 6.10 Relationships determined between shoot Zn concentration and soil Zn concentrations obtained using various methods; (A) Total Zn, (B) EDTA-extractable Zn, (C)
Ca(NO3)2 extractable Zn, (D) soil solution, (E) effective concentration, (F) free ion activity. A linear regression line is fitted for each plot. n = 594
141 6.4 Discussion
6.4.1 Effects of metal spiking of biosolids and soils on estimates of bioavailability
Pooled plant and soil data were subdivided into more specific datasets based on differences in biosolids and metal salt treatments. The purpose of this was to assess the degree of variability in shoot metal concentration that each test could significantly explain (P<0.05) in smaller more defined datasets. Variations in shoot metal concentrations described by each test in these smaller groups were also compared with correlation results from pooled plant and soil data.
There were marked differences in the relative abilities of the assays to account for shoot metal concentrations between the groups and metals. For Cd, bioavailability assays were unable to describe greater than 50% of variation in shoot concentrations. Results for Cu revealed that measures of bioavailability were also unable to describe greater than 50% of variation in shoot Cu conentrations, with the exception of soils treated with metal spiked biosolids. While measures of Ni and Zn bioavailability were able to explain greater than 50% shoot metal concentration, but only in groups containing metal salts.
Explanatory ranges for Cd for the separate groups were between 4% and 40% (Table 6.1) for all methods tested and were considerabley lower in range compared to pooled plant and soil data (16% to 47%, Figure 6.7). Effective Cd concentration in the pooled data could describe the greatest amount of variability in shoot Cd concentration
(47%). However when assessed in the smaller groups, CE-Cd was able to describe less than half of the variability in shoot Cd concentration (15% to 21%), and only in soil
spiked with Cd salts. In soils having received no biosolids or metal salts, Ca(NO3)2 and EDTA extractable Cd could describe between 19 % and 40% shoot Cd concentration. This range is closer to that obtained from the pooled plant and soil data
(30% and 33% for Ca(NO3)2 and EDTA extractable Cd, respectively). These results suggest that the presence of Cd as a metal contaminant strongly determines the descriptive range of bioavailability measures such as DGT. While in unamended soils methods such as Ca(NO3)2 and EDTA extractable Cd more accurately reflect plant available Cd.
142
Explanatory ranges for variation in shoot Cu concentrations were highest in soils amended with metal spiked biosolids and measures of Cu soil bioavailability were able to describe between 23% and 63% of shoot Cu concentration (Table 6.2). Soils spiked with Cu salts gave similar ranges (32% to 43%) for the six methods assessed, compared with the ranges determined for pooled plant and soil data (26% to 34%,
Figure 6.8). Of these tests EDTA extractable Cu and CE-Cu accounted for the greatest amount of shoot Cu concentration in soils amended with metal spiked biosolids (55% to 63%), which were higher than soils amended with Cu salts (30% to 35%) and also results for pooled plant and soil data (26% to 27%, Figure 6.8). Thus it appears that the presence of Cu as Cu salts strongly influences the explanatory range of these methods assessed. Moreover, metal salts incorporated into the biosolids matrix results in an increase in the bioavailability and plant uptake of Cu.
For Ni the degree of variability in shoot Ni concentrations described by the assays tested increased with the addition of Ni salts (Table 6.3). In soils spiked with Ni salts (biosolids and metals, and metals only), almost all of the methods assessed could account for greater than 50% of the variability in shoot Ni concentrations. However in soils that received metal spiked biosolids, only EDTA-extractable Ni was able to describe greater than 50% of variability in shoot Ni concentration. When the descriptive ranges of the smaller groups were compared to the relationships determined between pooled plant and soil data (Figure 6.5), soils that had been treated with Ni salts gave similar ranges to those obtained from pooled results. The presence of biosolids including the spiking of biosolids with metal salts appeared to decrease the descriptive power of the assays tested. Hence, these results suggest the presence of Ni as a metal salt in soils strongly influences the outcome of these bioavailability
assays with Ca(NO3)2-extractable Ni and soil solution Ni having the highest correlation with shoot Ni concentrations (between 61% and 81% for soil solution Ni
and Ca(NO3)2-extractable Ni, respectively).
The number of significant relationships for shoot Zn concentration and assays of Zn bioavailability were higher across the five groups compared to Cd, Cu and Ni (Table 6.4). However, the general trend of increasing descriptive range of these tests with the presence of Zn in soils as Zn salts was similar to that of the other three metals. Of the
143 groups compared, metal spiked biosolids gave correlations similar to those obtained from the pooled plant and soil data (Figure 6.6). Results from assays on soils treated with metal salts only, could explain a higher degree of variability in shoot Zn concentrations than from plants grown in metal spiked biosolids, as well as relationships determined using pooled plant and soil results. However, only in two of the groups (metal spiked biosolids and metal salts only) were any assays able to describe greater than 50% of variability in shoot Zn concentration, and of these tests
Ca(NO3)2-extractable Zn and CE-Zn were the most robust measures of Zn bioavailability (61% to 70%, Table 6.4). Thus it appears that the power of assays of bioavailability is influenced by the concentrations of Zn in biosolids with higher
concentrations of Zn showing greater descriptive ranges.
The ability of each of these methods to significantly predict metal concentrations in shoots appeared to be largely influenced by the concentration of metal present in the biosolids and soil (i.e. the more metal present the higher r2 value). The findings in this study suggest that three factors need to be considered when assessing the robustness of any method in estimating bioavailability:
• The range of variability that each assay of bioavailability can describe across a range of soils and soil amendments, i.e. an assay that presents a narrow range (e.g. 60% to 70%) may be more robust than an assay that provides a wider descriptive range (e.g. 40% to 75%);
• total metal concentration and likely form of metal (e.g. mostly bound to organic anions, or sorbed to organic matter) which would depend on the nature of the study, e.g. heavily contaminated soil, or assessment of naturally occurring elevated metal concentrations in serpentine soils;
• if the metal is known to accumulate around the roots or translocate to the shoots of the plants.
The appropriateness of methods used to estimate bioavailability have been discussed in a few recent studies (Basar, 2009; Degryse et al., 2009a; 2009b; McBride et al., 2009; Meers et al., 2007; Meers et al., 2009; Menzies et al., 2007; Nolan et al., 2005). McBride et al. (2009), concluded that chemically non-aggressive neutral salts, such as
CaCl2 may be the most appropriate extractants where phytotoxicity is the concern in
144 metal contaminated soils as these methods appear more responsive to soil properties affecting chemical lability of the metals. These findings certainly concur with the
findings of this chapter and that of Chapters 4 and 5, in that Ca(NO3)2 gave the strongest relationship across a range of soils and amendments, especially for Ni. Certainly methods that considered the response of the solid phase reservoir of metals
(i.e. Ca(NO3)2 and DGT) correlated better in general, and this has also been discussed in three previous studies (Degryse et al., 2009b; Lehto et al., 2006b; McBride et al., 2009). Although the predictive strength of these methods, especially DGT, is dependent on diffusive transport of the element from soil to the plant roots being rate- limited, competing cations are not affecting plant uptake of the metal in question and, and labile complexes are not taken up by the plant if diffusion is not limiting (Degryse et al., 2009b).
6.4.2 Effects of plant species on estimates of bioavailability
There are two major confounding factors in this assessment of plant species effects on estimates of bioavailability that may limit the interpretation of the results:
• plant maturity, i.e 20 day-old wheat seedlings versus mature rygrass;
• controlled environment versus field conditions.
In general assays of bioavailability were able to explain more variation in ryegrass shoot metal concentrations than wheat shoot metal concentrations. The relationships determined for both plant shoot metal concentrations and six measures of metal bioavailability were significantly different between species of plant for most of the assays tested. Overall shoot concentrations of Cd and Zn were significantly higher in wheat, while shoot concentrations of Ni was significantly higher in ryegrass. No assay was able to describe more than 50% variation in Cu shoot concentration data for either plant species.
Cadmium in soil solution and CE-Cd appeared to be a dominant source of plant available Cd for ryegrass grown under field conditions (Table 6.5). When the explanatory ranges for each plant species were compared to relationships obtained for
145 pooled plant and soil data for Cd (Figure 6.7), pooled correlation values were generally lower, or similar to that obtained for wheat.
This trend was also observed with Ni concentrations in ryegrass shoots in that
bioavailable Ni as measured by CE was the dominant source of plant available Ni compared with wheat grown in pots. However, wheat also showed similar relationships with assays of Ni bioavailability to ryegrass. Comparing the individual species relationships with assays of bioavailability to the overall relationship determined for pooled plant and soil data revealed that pooled coefficient of determination values (Figure 6.9) were more closely aligned with relationships determined for wheat.
Plant species relationships with Zn and assays of Zn bioavailability were the least different of all the metals assessed. Only pools of Zn as measured by EDTA and
Ca(NO3)2, were significantly different with wheat having an overall higher uptake of Zn from available pools as measured by these methods. Relationships determined between pooled plant and soil data and six measures of Zn bioavailability appear more closely aligned with correlation values obtained using wheat. However, the degree to which these tests can significantly describe shoot Cu and Zn variability for the two species of plant are not as robust or consistent as they are for Cd and Ni, which suggests that either plant (or trial) may adequately describe variation in shoot Zn concentration.
Differences in plant species uptake of metals has been documented in several studies (Almas et al., 2006; Chen et al., 2009; Dunbar et al., 2003; Gray and McLaren 2005; Hamon et al., 1997). For example, in a pot trial carried out by Almas et al. (2006) using two species of plants, ryegrass, (Lolium multiflorum), and spinach (Spinacia oleracea) grown in metal contaminated soils in a controlled environment, good 2 correlations were found between ryegrass Zn and CE-Zn as measured by DGT (r =
0.96 for Zn), while no significant correlation was found for CE-Zn and spinach Zn concentrations. Moreover, Gray and McLaren (2005) found that even within the ryegrass genus, there was a significant amount of intraspecies variation in herbage metal concentrations in plants grown in the same contaminated soil. Another successful relationship between plant Zn concentration and CE-Zn was also observed
in lettuce (Lactuva sativa) grown in contaminated agricultural soils, but a less robust
146 relationship was observed with Cd concentrations in the leaves and CE Cd as measured by DGT (Cornu and Denaix, 2006).
Results from this study concur with findings reported by Almas et al. ( 2006) and Cornu and Denaix (2006) which suggests that the strength of potential measures of metal availability is species dependent. Additionally, relationships between availability measurements and plant concentrations would likely be affected by plant health as metal uptake generally decreases in plants affected by metal toxicity, whereas most measures of metal solubility are not limited by a physiological response. Other factors to consider regarding the differences of relationships determined between plant metal concentrations and measures of metal availability are:
• sources of available metals other than the resupply of metals weakly bound to the solid phase of the soil, e.g. metals bound to low affinity ligands;
• metal species other than the free ion are being taken up by plants, e.g. CdCl+ (Degryse et al., 2006a; Degryse et al., 2006b; Weggler et al., 2004);
• competitive uptake of other divalent cations such as Ca, Mg (Kalis et al., 2006);
• the possibility that certain species may specifically exclude the uptake of some divalent cations, or internal detoxification by the plant, e.g. by binding Cd to phytochelatins, also the lack of translocation from the root to the shoot (Reid et al., 2003a);
• rhizosphere processes (including root morphology) that may alter the availability of the metal, such as the exudation of low molecular weight organic compounds combined with alteration of pH (Gregory, 2006; Jones et al., 2004; Lu et al., 2007).
The power of measures of bioavailability to predict shoot concentrations were significantly different for ryegrass and wheat. Overall assays were able to describe more variation in ryegrass shoot data than for wheat, with Ca(NO3)2, DGT and soil solution being the most successful measure of Cd, Ni and Zn concentration in
147 ryegrass shoots. However, when comparing relationships obtained from pooled plant and soil data, these were more closely aligned with relationships obtained with 20 day-old wheat seedlings. In summary, results suggest that the use of a wheat seedling assay may not adequately reflect metal bioavailability in longer term studies of metal contamination and plant uptake. Therefore the use of wheat seedling assays as model systems carried out in controlled environmental conditions to evaluate metal bioavailability may need to be standardised against mature plant models managed under field conditions.
6.4.3 Evaluation of six potential measurements of bioavailability using pooled plant species and soil data
In total, 12 soils were amended with different combinations and rates of biosolids and metal salts and two species of plant (wheat and ryegrass) were chosen as the biological model systems to evaluate the response of plants to soil amendments from studies carried out over several years. Although this comparison of pooled data from across several studies provided the opportunity to determine the robustness of these methods it also introduced several confounding factors that can potentially affect the predictive strength of these assays and the interpretation of results. Such factors were described in previous sections and included: field experiments versus controlled environment; wheat versus ryegrass responses, and mature plants versus 20 day old seedlings.
The range of predictive strengths for Cd bioavaibility found in this study, are also similar to those described previously in the literature. Studies assessing the predictive capabilities of methods (i.e. DGT, speciation models, free ion concentration, EDTA- extraction and weak salt solutions) to determine shoot Cd concentrations also found a range of coefficient determinations for Cd as measured by these methods (Almas, et al., 2006; Gray et al., 2004; Menzies, et al., 2007; Nolan, et al., 2005). The robustness of these potential measures of bioavailability appeared dependent on the supply of the metal (Almas, et al., 2006; Menzies, et al., 2007; Nolan, et al., 2005) and the plant species involved (Almas et al., 2006).
148 In this study, no method assessed could explain more than 50% of variability in shoot
Cd concentration. However, CE-Cd was able to explain the most variation in shoot Cd concentration (47%). The inability of any of the methods assessed to significantly predict greater than 50% of the variability for Cd could be due to the relatively low concentrations of Cd present in the soil with respect to other cations such as Ca, Cu, Mg, Ni and Zn. This situation could have lead to the competive uptake of cations present in higher concentrations via non-specific cation pathways found in plant roots thereby limiting uptake of cations present in much lower concentrations, such as Cd (Reid et al., 2003a). Another factor that may have affected the ability of methods such as DGT to explain greater than 50% of variability in shoot Cd concentration are labile complexes of Cd not accounted for by DGT that may be taken up by plants (Degryse et al., 2009b). Moreover, unlike Cu, Ni and Zn, Cd is a non-essential element (Alloway, 1995a; Kiekens and Cottenie, 1985; Sloan et al., 1997) and may be actively excluded from shoots by the plant, and thus affect the amount of shoot Cd concentration variation described by the methods assessed.
Several studies assessing the predictive power of assays of metal bioavailability have found that predictions of Cu uptake by plants from soils by the various measures of available Cu are typically poor (r2 < 0.5) (Basar 2009; Nolan et al., 2005; Menzies et al., 2007), except for total Cu (Nolan et al., 2005; Menzies et al., 2007), which is in agreement with the predictive results obtained in study. This result is unsurprising as plants typically exert strong physiological controls on the translocation of Cu from the roots to the shoots (Reid et al., 2003a). Zhang et al. (2001) were able to predict 95% of the variability in plant Cu using DGT in a pot trial using 29 soils and Smith’s pepperwort (Lepidium heterophyllum) as the biological response model. However, the soils used in the study were contaminated mostly with Cu and the concentrations present in the soils were more than 8 times higher (> 8000 mg kg-1) than Cu concentrations present in any of the treated soils used in this study (highest treatment application was 1000 mg kg-1). In addition, L. heterophyllum is also known to be an effective accumulator of Cu (Nolan et al., 2005) and thus does not tightly regulate Cu translocation from the roots to the shoots as in non-accumulating plant species.
Bioavailability measures of Ni and Zn were the most successful of the four metals in
determining shoot metal concetrations. The validity of 0.05 M Ca(NO3)2 extraction as a reasonable predictive measure of Ni and Zn bioavailability is evident given that
149 these results were obtained from a range of New Zealand soils, biosolids and metal concentrations and two plant species. The success of neutral salt extractants such as
such as 0.05 M of Ca(NO3)2 as an indicator of metal availability has been briefly described by Menzies et al. (2007) and Meers et al. (2009) These extractants were developed on the theory that bioavailable metals are located on soil mineral surfaces and these can be displaced by competing cations (Andrews et al., 1996; McLaughlin et al., 1999).
6.5 Conclusions
The predictive powers of each bioavailability method used in this study rely on the theory that the concentration of metal aquired will reasonably reflect the concentrations found in plant shoots. Although there was a considerable amount of
variation in correlation strengths of Ca(NO3)2-extractable metals, this method significantly accounted for the most variation in shoot metal concentrations observed for Ni (74%) and Zn (65%). Total metal, EDTA extraction and DGT had the lowest explanatory range (33 – 34%) and were able to explain greater than 50% of the shoot concentration variation for Ni and Zn, but were on average lower in their predictive
strength for Ni and Zn than for Ca(NO3)2-extractable Ni and Zn. Correlation results for Cd and especially Cu were weak and none of the methods assessed could explain greater than 50 % of variability, with the lowest predictive capability obtained using Cd2+ activity (r2 = 0.16).
The responses of wheat and ryegrass as assessed using six potential measures of bioavailability were consistently different. Concentrations of Cd and Zn were generally higher in wheat than in ryegrass, while concentrations of Ni were typically higher in ryegrass. Apart from Cu, the methods assessed were able to significantly explain a greater amount of variability for ryegrass data than for wheat, which suggests a greater predictive capability for assessments involving ryegrass. Unsurprising, shoot concentrations of each metal were different for wheat and ryegrass. Concentrations of metals in wheat shoots were generally higher in relation to larger sized pools of available metals as described by EDTA, Ca(NO3)2 and soil
150 solution, whereas ryegrass shoot concentrations of metals were generally higher in relation to the much smaller pools of exchangeable metals, as measured by DGT. Hence this suggests that longer growth periods had a greater demand on the soil system.
Interpreting the relationships between shoot metal concentrations and assays of metal bioavailability in five soil treatment groups that share similarities for the purpose of assessing method validity is complicated. However the findings of this study suggest that two factors need to be considered when assessing a method’s robustness and appropriateness in estimating bioavailability:
• range of r2 values for the method across a range of soils and soil amendment;
• the total soil concentration of the metal and whether that metal is known to accumulate, or translocate to the tops of the plants.
Methods that consider the biologically relevant amount of metal supplied from the solid phase, appear to provide a more robust measure of bioavailability and this study had proven this to be the case for Zn and especially Ni, where the validity of these methods were determined using two species, soil types and amendment regimes.
151 Chapter 7 General Discussion, Conclusions and Approaches for Further Studies
7.1 Summary
It is widely accepted that metal bioavailability is important in providing accurate environmental risk assessments of metal contaminated soils. However there is still uncertainty regarding mechanisms that control the pathway of metals from the soil to the biological endpoint (i.e. leafy shoots) (Palmer and Guerinot 2009; Reid et al., 2003a). Commonly, the concentration of metal considered bioavailable is often defined as the fraction of metal extracted from soil using a chemical reagent, based on correlation with plant shoot metal concentrations (Menzies et al., 2007). In most cases bioavailability is largely a measure of metal solubility. Existing techniques have been developed and/or modified for use in soils, which have attempted to provide a more inclusive or, targeted approach in determining bioavailability. These include:
• the development and modification of the DGT method which simulates plant uptake of metals through diffusion via the roots (Lehto et al., 2006a; Zhang et al., 2001). The DGT method is used in conjunction with the DIFS modelling programme that considers the replenishment capacity of the solid phase to resupply the amount depleted in soil solution (Harper et al., 2000);
• assemblage models that determine free ion activity and speciation in soil solution (WHAM) (Degryse et al., 2009a; Tipping et al., 2003b);
• neutral salt solutions with ionic strengths similar to that of soil solution that saturate the system with competing cations, thereby displacing the fraction of metal considered readily available for plant uptake (Andrews et al., 1996; Meers et al., 2009; Menzies et al., 2007).
152 However, there are insufficient datasets that span confounding factors such as soil properties, metal concentrations, soil amendments and plant species to rigorously verify these methods.
The overall aim of this thesis was to assess and validate measures of bioavailability of Cd, Cu, Ni and Zn across various soils amended with metals and biosolids. This was achieved by assessing six potential measures of bioavailability in three distinct, but complimentary chapters:
• Chapter 4 assessed the effects of biosolids amendment on the bioavailability of Cd, Cu, Ni and Zn in three metal spiked soils under controlled environmental conditions, using wheat as the model plant system;
• Chapter 5 assessed the effects of biosolids and metal salt amendment on the bioavailability of Cd, Cu, Ni in the same three soils used in Chapter 4; managed under field conditions, using ryegrass as the model plant system;
• Chapter 6 assessed the overall strength of the relationships between shoot metal concentrations in wheat and ryegrass and potential measures of Cd, Cu, Ni and Zn bioavailability from 12 New Zealand soils, including those soils used in Chapters 4 and 5, amended with different combinations of biosolids and metal salts.
7.2 Estimates of bioavailability
A total of 12 soils were amended with different combinations of biosolids and metal salts. Two species of plant (wheat and ryegrass) were chosen as the biological model systems to evaluate the response of plants to treated soils from data gathered between 2002 and 2008.
Of the four metals examined, Ni was found to have the strongest relationship between measures of bioavailability and shoot concentrations (Table 7.1). Results indicated that shoot concentrations of Ni were highly indicative of the concentration of Ni present in soil solution and Ni weakly sorbed to soil particles. Of the six methods
153 assessed, Ca(NO3)2 was the most robust assay and could predict 73% to 87% of the variability in Ni concentrations in the shoots of both species. However, all methods assessed were relatively good predictors of bioavailable Ni and could explain greater than 50% of shoot responses (Table 7.1).
154 Table 7.1 Range of significant (P<0.05) coefficients of determination for six potential measures of bioavailability assessed in three related studies. Total varation for each method and metal is derived from the highest r2 value minus the lowest r2 value.
Chapter 4: Pot trials using wheat Total Method Cd Cu Ni Zn variation (method) Total metal 60% 15% 63% 38% 48% EDTA 63% 13% 60% 54% 50%
Ca(NO3)2 62% 10% 73% 55% 63% Soil solution 30% 6% 66% 33% 60% DGT 26% 13% 44% 43% 30% Free ion activity 30% 11% 67% 33% 56% Total variation 37% 9% 29% 22% (metal) Chapter 5: Field lysimeter trial using ryegrass Total Method Cd Cu Ni Zn variation (method) Total metal 73% 49% 82% 55% 33% EDTA 28% 17% 44% 35% 27%
Ca(NO3)2 34% 59% 87% 43% 53% Soil solution 71% 49% 83% 68% 34% DGT 65% 25% 84% 76% 59% Free ion activity 64% 7% 75% 68% 68% Total variation 45% 52% 43% 41% (metal) Chapter 6: Combined Lincoln University and ESR trials Total Method Cd Cu Ni Zn variation (method) Total metal 42% 34% 67% 52% 33% EDTA 33% 22% 56% 52% 34%
Ca(NO3)2 30% 30% 74% 65% 44% Soil solution 19% 29% 64% 56% 45% DGT 47% 26% 54% 60% 34% Free ion activity 16% 30% 55% 52% 39% Total variation 31% 12% 18% 13% (metal)
155 The robustness of Ca(NO3)2 extraction for estimating Zn bioavailability was unable to describe the same degree of variability in shoot metal concentration as was shown for Ni. However, results from the pot trial did give a higher degree of predictive power than results from the field trial (55% versus 43% for each respectively, Table 7.1). Although in Chapter 6, where ESR studies were combined with results presented in
Chapters 4 and 5, the predictive power of Ca(NO3)2-extractable Zn increased to 65%. The overall strengths of the six measures of bioavailability to predicted Zn concentration in shoots were stronger in the field lysimeter trials than in the pot trials, 2+ with soil solution Zn, CE-Zn and Zn activity significantly explaining 68% to 76% of the variations in shoot Zn concentrations (Table 7.1). The difference between the experiments in ability to predict bioavailable Zn most likely reflects three major confounding factors:
• field studies versus controlled environments (i.e. leaching of cations and anions through the soil profile);
• wheat versus ryegrass;
• 20 day old seedlings versus mature plants.
These factors can alter the effect that the presence of competitive cations, or complexes, that these tests that may overcompensate for, or complexes that are (un)available for plant uptake (Almas et al., 2006; Degryse et al., 2006b; 2009b). For example, the trials in pots were not free draining and hence, would retain many of the cations and anions (both inorganic and organic) that were lost as a result of drainage that occurred in the field trial. The retention of soil solution in pots could also allow a greater amount of interaction between components in the solution and exchange with the solid phase of the soil. Also, seedlings have different nutritional requirements to more mature plants and in longer term trials, the most available fraction (i.e. soil solution) would have been depleted within the first few months.
The results for Cd were generally similar to Zn and were explain very similar percentages of variation (50.5% versus 50.1% for Cd and Zn respectively). Additionally results from the lysimeter study could explain the greatest amounts of shoot Cd variation compared to Cd results from pot trials (Table 7.1). However, results from the combined studies (Chapter 6) revealed an overall decrease in the
156 predictive capabilities of all methods assessed compared to results from field trials, with no assay being able to describe more than 50% of shoot Cd variability (Figure 6.7). The lower predictive power of the soils tests to describe Cd variability in Chapter 6 compared with Chapters 4 and 5 is most likely due to the Cd data mostly being from pot trials where ranges of Cd concentrations in the soil were low compared with other metals. Comparitively low soil concentrations of Cd compared with Cu, Ni and Zn would result in greater concentrations of those metals present in higher concentrations being taken up by plants, because of the effect of mass action. This result also suggests that competing cations such as Ca and Zn may also affect the outcome of these tests with respect to Cd as these would be present in much higher concentrations.
None of the methods assessed could explain more than 50% of shoot Cu variability, although total Cu and soil solution Cu results from the field lysimeter trial did result in correlation of r2 = 0.49 for both methods. Although some studies have shown good fits with potential measures of bioavailability and Cu uptake by plants (Zhang et al., 2001), the majority of research has found poor relationships between potential bioavailability measure and shoot Cu concentrations (Menzies et al., 2007). The reason for this is that Cu in soil solution is strongly complexed by organic matter, hence most Cu is considered unavailable for plant uptake, and the uptake of Cu to shoots from roots is strongly regulated by plant metabolism of nonaccumulating plants.
The overall efficacy of the different methods for assessing bioavailability appeared to be largely dependent on the metal itself (i.e. Ni was a good candidate for solubility based tests, while Cu was not), the extent of metal contamination (higher and greater range of concentrations of metal in the soil gave stronger relationships with shoot metal concentrations), the presence of competitive cations and complexing anions (i.e. Ca, Mg competing with Cd and Zn for uptake, and DOC complexing with Cu and Ni making Cu and Ni less/more available for plant uptake), and the general condition of the plant.
157 7.3 Influence of biosolids on metal bioavailability
Chapters 4 (pot trial) and 5 (field trial) explored the potential effects that biosolids amendment might have on the bioavailability of Cd, Cu, Ni and Zn, In the pot trial using wheat, all metal salt treatments had corresponding biosolids amendments, while in the field trial using ryegrass, only the lowest concentration of metal salt treatments was amended with biosolids. For both studies the addition of biosolids had a marked effect on soil chemistry, with increases in DOC, Ca, Mg and salinity observed in the soil solution. Also in the pot trial, an increase in total soil Cu and Zn was observed, largely owing to the ubiquitous presence of Cu and Zn in biosolids. Increased DOC, Ca, Mg and salinity were correlated with increased soil incubation time (pot trial) and length of time since treatment application (field trial).
In the pot trial, the presence of biosolids did not significantly affect the extractability or solubility of any of the metals as measured by five of the soil tests (EDTA, Ca(NO3)2, soil solution, DGT and free ion activity, Table 4.4). The relationship determined between shoot metal concentrations and soil assasys of metal bioavailability were also not affected by the presence of biosolids. This result suggests that the presence of biosolids did not increase the potential bioavailability of the four metals.
Results from the field trial showed that in the presence of biosolids increased EDTA- 2+ extractable Cd, Cu activity, EDTA-extractable Ni and CE-Zn, but decreased CE-Cd, CE-Ni, Ni2+ activity and Zn2+ activity (Table 5.5). The relationships determined between shoot Cd and Ni concentrations and assays of Cd and Ni bioavailability were also significantly different in the presence of biosolids (Table 5.16 and 5.18). Three possible explanations could account for these results:
• influence of time on the system (six months versus 24 months) resulted in increased sorption of Cd and Ni into the biosolids matrix as observed by the increase in EDTA-extractable Cd and Ni, and the decrease in the readily
available fraction of Cd and Ni (CE and free ion activity);
• greater plant demand on the system as a result of longer growth period (e.g. the removal of readily available Zn), with the presence of biosolids providing an additional supply of Zn to replenish the amount of Zn in solution (increase in
CE-Zn);
158 • leaching of cations (Ca and Zn), anions and DOC through the profile would remove ions and DOC from the top 10 cm of the soil profile, leading to a decrease in ionic strength of the soil solution that may result in an increase in Cu2+ activity.
7.4 Plant responses
In the pot trial, a significant reduction in dry weight was recorded in soils amended with biosolids. While several factors, including the toxic effects of metals were explored, results revealed that salinity of the soil and to a lesser extent Cu and Ni concentrations within the shoots was controlling the production of dry matter (Figure 4.9, 4.12 and 4.13). It was expected that symptoms of phytotoxcity would be exhibited by plants grown in the highest metal treatments, although the osmotic stress induced by the presence of biosolids was unexpected, with the dissolution of metal salts having also contributed to the overall osmotic stress experienced by the plants. Changes in dry weight were not determined in the lysimeter study, as only a representative plant sample was collected at each harvest and so comparisons were not made between the pot and field trials.
The presence of biosolids resulted in a decrease in Cd concentration in wheat in the pot trial (Table 4.6 and 4.7). However this effect was only observed at the 24 month harvest period in the lysimeter trial (Table 5.14). Although the significance level for the 18 month harvest period in the field trial was only just greater than 0.05 (P = 0.057). Even though Cd results from the lysimeter trial were less conclusive than from the pot trial, inferences based on soil chemistry results from both trials suggest that the presence of biosolids decreases plant available Cd, and from these two mechanisms may be responsible for this effect:
• increased concentration of DOC (Table 4.2 and 5.2) reducing the availability of Cd by complexation;
• presence of cations such as Ca and Mg(Table 4.2 and 5.2) that would competitively inhibit the uptake of Cd by plants.
In the pot trial, increased soil incubation time resulted in the significant decrease in Cu concentration in wheat, while generally Zn concentration in wheat increased significantly (Table 4.6). The same result for Zn concentration in ryegrass was also observed in the field trial, while overall Ni concentration in ryegrass decreased with time in the field trial. It was
159 inferred from soil chemistry results that complexation by soil organic matter (Cu) and DOC (Cd and Ni), and an increased demand from the plant system (Zn) with increased availability (in the case of the pot trial) were responsible for these observed trends.
Chapter 6 compared the two plant species at different stages of growth and their relationship to potential measures of bioavailability. The relationships of wheat and ryegrass were consistently significantly different from each other. Overall concentrations of Cd and Zn were significantly higher in wheat than in ryegrass, while Ni was present in significantly higher concentrations in ryegrass. In relation to Cu responses, ryegrass had significantly lower shoot concentrations of Cu than wheat as determined by relationships between Ca(NO3)2, soil solution Cu and CE-Cu and shoot Cu, although regression values for Cu were typically poor (r2 <0.45). Coefficients of determinations between shoot concentrations and potential measures of metal bioavailability were higher for ryegrass than for wheat. In addition, results suggested that metal concentrations measured by 0.05 M of Ca(NO3)2, extraction, CE and metals in soil solution were the dominant pools of available Cd and Ni for ryegrass, while available pools of Zn as measured by EDTA and Ca(NO3)2 were a dominant source of Zn for wheat grown in pots compared to than for ryegrass. This suggests that longer growth periods (i.e. ryegrass in the field trial) deplete the most readily available sources of metals, which then result in these metals having to be replenished from the solid phase.
7.5 Conclusions
There are many extractive techniques used to determine the bioavailable metal fraction in soils but currently there is no consensus as to the most appropriate measure of metal bioavailability. The predictive power that these methods have are based on hypotheses that the amount of metal present in solution along with varying amounts of metal extracted from the solid phase of the soil will reasonably reflect concentrations in the above ground plant tissues. Investigation of six potential measures of bioavailability across soils with contrasting properties, range of metal concentrations, presence of biosolids and two plant species carried out in this study represented a rigorous assessment of this topic. The degree to which these tests could signficantly explain the variability in shoot metal concentrations appeared to be metal specific, with the concentration of metal in the soil, competing cations and the presence of DOC influencing the availability. Of these tests Ca(NO3)2 provided the most robust measure of Cd, Zn and especially Ni bioavailability with results for Ni able to account for
160 more than 80% of the variability in shoot Ni concentrations in the lysimeter field trial. Thus
Ca(NO3)2 provided the most useful indication of bioavailability across soils, soil amendments and plant species for Cd, Ni and Zn.
Results for other measures of Ni availability yielded the strongest correlations overall, suggesting that the solubility of Ni is highly indicative of shoot concentrations. On the other hand, results for measures of bioavailable Cd, Zn and especially Cu were less indicative of shoot concentrations. Relationships between shoot concentrations of Cd and Zn and corresponding soil concentrations require further investigations. Results for Cu were expected as assays of Cu solubility are almost always poorly correlated with shoot concentrations, hence tests based on the extractability and solubility of metals are not robust measures of predicting phytotoxic concentrations of Cu in soils. The addition of biosolids did not alter the outcome of these bioavailability tests in the pot trials presented in Chapter 4 and results presented in Chapter 6 suggest that it is the range of metal concentrations present in the soil and biosolids that have a greater influence on r2 values, rather than simply the presence of biosolids. Findings presented in Chapter 4 also indicated that the presence of biosolids in a particular treatment appeared to decrease the concentration of Cd in shoots, although the exact mechanism(s) responsible were inferred but not determined. While species differ in their uptake of metals from soils, results from this study suggest that both species access the same bioavailable pool of metals. Also the disparity between wheat and ryegrass is highly likely to be reflective of the different growth periods (i.e. 20 days versus 24 months), which would exert different demands on the system. Hence standardising seedling assays against mature plant models may be necessary for bioavailability assessments carried out on longer term studies of metal contamination.
7.6 Approaches for further studies
The findings of the research presented in this thesis suggest further studies in the following areas:
• identification and quantification of organic anions excreted by the plant roots – are some more prevalent than others? and do certain organic anions have a greater affinity for metals than others? Would these alter the bioavailability of Cd, Cu and Zn? By quantifying these organic anions and their affinity for metals improve the
161 strength and consistency of the relationship between soil measures of Cd, Cu and Zn and plant concentrations?;
• the use of root Cu concentration instead of shoot Cu concentration as a biological endpoint.
• does root morphology impact upon metal uptake? Different species of plants have different root systems, would this alter metal uptake?;
• does soil microbial activity affect metal bioavailability? if so are there specific communities associated with plant species, especially commercial crops, that can be identified?;
• do mychorrizae species affect metal bioavailability during their acquisition of P from soil minerals?
162 References
Adams M.L., Zhao F.J., McGrath S.P., Nicholson F.A., Sinclair A.H. (2001) Trace elements in UK wheat: from selenium deficiency to heavy metal contamination by cadmium and lead. Aspects of Applied Biology:209-216. Allen H.E., McGrath S.P., McLaughlin M.J., Peijnenburg W.J.G.M., Sauve S., Lee C. (2001) Bioavailability of metals in terrestrial ecosystems: importance of partitioning for bioavailability to invertebrates, microbes, and plants. Bioavailability of metals in terrestrial ecosystems: importance of partitioning for bioavailability to invertebrates, microbes and plants:xi + 158 pp. Alloway B.J. (1995a) Heavy metals in soils. Heavy metals in soils.:xiv + 368 pp. Alloway B.J. (1995b) Toxic metals in soil-plant systems: Edited by Sheila M. Ross. John Wiley and Sons, Chichester, UK, 1994, ISBN 0-471-94278-0, 469 pp. Price: £55.00 (hardback). Environmental Pollution 90:270-270. Alloway B.J. (2009) Soil factors associated with zinc deficiency in crops and humans, Symposium on Soils and their Implication on Health, Vienna, AUSTRIA. pp. 537-548. Alloway B.J., Jackson A.P. (1991) The behaviour of heavy metals in sewage sludge-amended soils. Science of the Total Environment 100:151-176. Alloway B.J., Thornton I., Smart G.A., Sherlock J.C., Quinn M.J. (1988) Metal availability. Science of the Total Environment 75:41-69. Alloway B.J., Baker, D.E., Davies, B.E. (1990) Heavy metals in soils Blackie & Son Ltd, Glasgow. Almas A.R., Bakken L.R., Mulder J. (2004) Changes in tolerance of soil microbial communities in Zn and Cd contaminated soils. Soil Biology and Biochemistry 36:805- 813. Almas A.R., Lombnaes P., Sogn T.A., Mulder J. (2006) Speciation of Cd and Zn in contaminated soils assessed by DGT-DIFS, and WHAM/Model VI in relation to uptake by spinach and ryegrass. Chemosphere 62:1647-1655. Almas A.R., Lofts S., Mulder J., Tipping E. (2007) Solubility of major cations and Cu, Zn and Cd in soil extracts of some contaminated agricultural soils near a zinc smelter in Norway: modelling with a multisurface extension of WHAM. European Journal of Soil Science 58:1074-1086. Amery F., Degryse F., Degeling W., Smolders E., Merckx R. (2007) The copper-mobilizing- potential of dissolved organic matter in soils varies 10-fold depending on soil incubation and extraction procedures. Environmental Science and Technology 41:2277- 2281. Amery F., Degryse F., Cheyns K., De Troyer I., Mertens J., Merckx R., Smolders E. (2008) The UV-absorbance of dissolved organic matter predicts the fivefold variation in its affinity for mobilizing Cu in an agricultural soil horizon. European Journal of Soil Science 59:1087-1095.
163 An Y.J. (2004) Soil ecotoxicity assessment using cadmium sensitive plants. Environmental Pollution 127:21-26. Andrews P., Town R.M., Hedley M.J., Loganathan P. (1996) Measurement of plant-available cadmium in New Zealand soils. Australian Journal of Soil Research 34:441-452. Antoniadis V. (2008) Sewage sludge application and soil properties effects on short-term zinc leaching in soil columns. Water Air and Soil Pollution 190:35-43. Antoniadis, V., & Alloway, B. J. (2001). Availability of Cd, Ni and Zn to ryegrass in sewage sludge-treated soils at different temperatures. Water Air and Soil Pollution, 132(3-4), 201-214. Antoniadis V., Alloway B.J. (2002a) Leaching of cadmium, nickel, and zinc down the profile of sewage sludge-treated soil. Communications in Soil Science and Plant Analysis 33:273-286. Antoniadis V., Alloway B.J. (2002b) The role of dissolved organic carbon in the mobility of Cd, Ni and Zn in sewage sludge-amended soils. Environmental Pollution 117:515-521. Antoniadis V., Alloway B.J. (2003a) Evidence of heavy metal movement down the profile of a heavily-sludged soil. Communications in Soil Science and Plant Analysis 34:1225- 1231. Antoniadis V., Alloway B.J. (2003b) Influence of time on the plant availability of Cd, Ni and Zn after sewage sludge has been applied to soils. Agrochimica 47:81-93. Antonladis V., Tsadilas C.D. (2007) Sorption of cadmium, nickel, and zinc in mono- and multimetal systems. Applied Geochemistry 22:2375-2380. Antonladis V., Robinson J.S., Alloway B.J. (2008) Effects of short-term pH fluctuations on cadmium, nickel, lead, and zinc availability to ryegrass in a sewage sludge-amended field. Chemosphere 71:759-764. Antunes P.M.C., Kreager N.J. (2009) Development of the terrestrial biotic ligand model for predicting nickel toxicity to barley (hordeum vulgare): ion effects at low pH. Environmental Toxicology and Chemistry 28:1704-1710. Antunes P.M.C., Hale B.A., Ryan A.C. (2007) Toxicity versus accumulation for barley plants exposed to copper in the presence of metal buffers: Progress towards development of a terrestrial biotic ligand model. Environmental Toxicology and Chemistry 26:2282-2289. Antunes P.M.C., Berkelaar E.J., Boyle D., Hale B.A., Hendershot W., Voigt A. (2006) The biotic ligand model for plants and metals: Technical challenges for field application. Environmental Toxicology and Chemistry 25:875-882. Arnold R.E., Hodson M.E., Comber S. (2007) Does speciation impact on Cu uptake by, and toxicity to, the earthworm Eisenia fetida? European Journal of Soil Biology 43:S230- S232. Ashworth D.J., Alloway B.J. (2004) Soil mobility of sewage sludge-derived dissolved organic matter, copper, nickel and zinc. Environmental Pollution 127:137-144. Ashworth D.J., Alloway B.J. (2007) Complexation of copper by sewage sludge-derived dissolved organic matter: Effects on soil sorption behaviour and plant uptake. Water Air and Soil Pollution 182:187-196.
164 Basar H. (2009) Methods for Estimating Phytoavailable Metals in Soils. Communications in Soil Science and Plant Analysis 40:1087-1105. Basta N.T., Gradwohl R., Snethen K.L., Schroder J.L. (2001) Chemical immobilization of lead, zinc, and cadmium in smelter-contaminated soils using biosolids and rock phosphate. Journal of Environmental Quality 30:1222-1230. Berkelaar E.J., Hale B.A. (2003) Cadmium accumulation by durum wheat roots in ligand- buffered hydroponic culture: uptake of Cd-ligand complexes or enhanced diffusion? Canadian Journal of Botany-Revue Canadienne De Botanique 81:755-763. Blakemore L.C., Searle, P.L., Daly, B.K. (1987) Methods for chemical analysis of soils, New Zealand Soil Bureau Scientific Report (80). pp. 9-12. Bolan N.S., Duraisamy V.P. (2003) Role of inorganic and organic soil amendments on immobilisation and phytoavailability of heavy metals: a review involving specific case studies. Australian Journal of Soil Research 41:533-555. Bolan N.S., Khan M.A., Donaldson J., Adriano D.C., Matthew C. (2003) Distribution and bioavailability of copper in farm effluent. Science of the Total Environment 309:225- 236. Broadley M.R., White P.J., Hammond J.P., Zelko I., Lux A. (2007) Zinc in plants. New Phytologist 173:677-702. Brown S., Christensen B., Lombi E., McLaughlin M., McGrath S., Colpaert J., Vangronsveld J. (2005) An inter-laboratory study to test the ability of amendments to reduce the availability of Cd, Pb, and Zn in situ. Environmental Pollution 138:34-45. Brown S.L., Compton H., Basta N. (2007) Field test ofln situ soil amendments at the tar creek national priorities list superfund site. Journal of Environmental Quality 36:1627-1634. Buekers J., Degryse F., Maes A., Smolders E. (2008) Modelling the effects of ageing on Cd, Zn, Ni and Cu solubility in soils using an assemblage model. European Journal of Soil Science 59:1160-1170. Burton E.D., Phillips I.R., Hawker D.W., Lamb D.T. (2005a) Copper behaviour in a Podosol. 2. Sorption reversibility, geochemical partitioning, and column leaching. Australian Journal of Soil Research 43:503-513. Burton E.D., Phillips I.R., Hawker D.W., Lamb D.T. (2005b) Copper behaviour in a Podosol. 1. pH-dependent sorption-desorption, sorption isotherm analysis, and aqueous speciation modelling. Australian Journal of Soil Research 43:491-501. Butterly C. Tang, C, Baldock, J. (2008) Effect of initial soil pH and organic matter type on soil pH change. Abstract presented at the Australasian Soil Science Society Conference, 1-5th December 2008, Palmerston North, New Zealand. Cameron K.C.S., M.P. McLay, C.D.A. Fraser, P.A. McPherson, R.J. Harrison, D.F. Harbottle, P. (1992) Lysimeters without edge flow: an improved design and sampling procedure. Soil Science Society of America Journal:1625-1628. Cances B., Ponthieu M., Castrec-Rouelle M., Aubry E., Benedetti M.F. (2003) Metal ions speciation in a soil and its solution: experimental data and model results. Geoderma 113:341-355. Carrillo-Gonzalez R., Simunek J., Sauve S., Adriano D. (2006) Mechanisms and pathways of trace element mobility in soils. Advances in Agronomy, Vol 91 91:111-178.
165 Cattani I., Fragoulis G., Boccelli R., Capri E. (2006) Copper bioavailability in the rhizosphere of maize (Zea mays L.) grown in two Italian soils. Chemosphere 64:1972-1979. Cavanagh J. Coakley., J. (2005) An introduction to policy on metal contaminants in: Moore, T. A. Black, A. Centeno, J.A. Harding J.S. Trumm, D.A. (Ed.), Metal Contaminants in New Zealand, Resolutionz Press, Christchurch. pp. 5-23. Chaignon V., Sanchez-Neira I., Herrmann P., Jaillard B., Hinsinger P. (2003) Copper bioavailability and extractability as related to chemical properties of contaminated soils from a vine-growing area. Environmental Pollution 123:229-238. Chaney R.L., Oliver D.P. (1996) Sources, potential adverse effects and remediation of agricultural soil contaminants. Contaminants and the Soil Environment in the Australasia-Pacific Region:323-359. Chaney R.L., Chen K.Y., Li Y.M., Angle J.S., Baker A.J.M. (2008) Effects of calcium on nickel tolerance and accumulation in Alyssum species and cabbage grown in nutrient solution. Plant and Soil 311:131-140. Chaney R.L., Malik M., Li Y.M., Brown S.L., Brewer E.P., Angle J.S., Baker A.J.M. (1997) Phytoremediation of soil metals. Current Opinion in Biotechnology 8:279-284. Chaney, R. L. Yin-Ming Li, S.L.B., Angle, S. Stuczynski, T.I. Daniels, W.L. Henry, C.L. Siebielec, G. Malik, M. Ryan, J.A. Compton, H. ( 2000) Progress in Risk Assessment for Soil Metals, and In-situ Remediation and Phytoextraction of Metals from Hazardous Contaminated Soils, US-EPA Phytoremediation: State of the Science Conference, Boston, MA. pp. 1 -22. Chaperon S., Sauve S. (2007) Toxicity interaction of metals (Ag, Cu, Hg, Zn) to urease and dehydrogenase activities in soils. Soil Biology and Biochemistry 39:2329-2338. Chapman P.M. (2008) Environmental risks of inorganic metals and metalloids: A continuing, evolving scientific odyssey. Human and Ecological Risk Assessment 14:5-40. Chaudri A., McGrath S., Gibbs P., Chambers B., Carlton-Smith C., Bacon J., Campbell C., Aitken M. (2008) Population size of indigenous Rhizobium leguminosarum biovar trifolii in long-term field experiments with sewage sludge cake, metal-amended liquid sludge or metal salts: Effects of zinc, copper and cadmium. Soil Biology and Biochemistry 40:1670-1680. Chen Z.H., Setagawa M., Kang Y.M., Sakurai K., Aikawa Y., Iwasaki K. (2009) Zinc and cadmium uptake from a metalliferous soil by a mixed culture of Athyrium yokoscense and Arabis flagellosa. Soil Science and Plant Nutrition 55:315-324. Chojnacka K., Chojnacki A., Gorecka H., Gorecki H. (2005) Bioavailability of heavy metals from polluted soils to plants. Science of the Total Environment 337:175-182. Cloutier-Hurteau B., Sauve S., Courchesne F. (2008) Influence of microorganisms on Cu speciation in the rhizosphere of forest soils. Soil Biology and Biochemistry 40:2441- 2451. Collins R.N., Merrington G., McLaughlin M.J., Morel J.L. (2003a) Transformation and fixation of Zn in two polluted soils by changes of pH and organic ligands. Australian Journal of Soil Research 41:905-917.
166 Collins R.N., Merrington G., McLaughlin M.J., Morel J.L. (2003b) Organic ligand and pH effects on isotopically exchangeable cadmium in polluted soils. Soil Science Society of America Journal 67:112-121. Connolly E.L., Guerinot M.L. (2002) Iron stress in plants. Genome Biology 3:1024. Cornu J.Y., Denaix L. (2006) Prediction of zinc and cadmium phytoavailability within a contaminated agricultural site using DGT. Environmental Chemistry 3:61-64. Courchesne F., Cloutier-Hurteau B., Turmel M.C. (2008) Relevance of rhizosphere research to the ecological risk assessment of trace metals in soils. Human and Ecological Risk Assessment 14:54-72. Cox J.E. (1978) Soils and agriculture of Part Paparua County, Canterbury, New Zealand., Soil Bureau Bulletin 34. New Zealand Department of Scientific and Industrial Research, Wellington. pp.128. Datta S.P., Young S.D. (2005) Predicting metal uptake and risk to the human food chain from leaf vegetables grown on soils amended by long-term application of sewage sludge. Water Air and Soil Pollution 163:119-136. Davison W., Zhang H. (1994) In-situ speciation measurements of trace components in natural-waters using thin-film gels. Nature 367:546-548. Davison W., Hooda P.S., Zhang H., Edwards A.C. (1999) DGT measured fluxes as surrogates for uptake of metals by plants (Reprinted from Advances in Environmental Research, vol 3, pg 550-555, 2000). Advances in Environmental Research 3:U14-555. Degryse F., Smolders E. (2006) Mobility of Cd and Zn in polluted and unpolluted Spodosols. European Journal of Soil Science 57:122-133. Degryse F., Smolders E., Merckx R. (2006a) Labile Cd complexes increase Cd availability to plants. Environmental Science and Technology 40:830-836. Degryse F., Smolders E., Parker D.R. (2006b) Metal complexes increase uptake of Zn and Cu by plants: implications for uptake and deficiency studies in chelator-buffered solutions. Plant and Soil 289:171-185. Degryse F., Verma V.K., Smolders E. (2008) Mobilization of Cu and Zn by root exudates of dicotyledonous plants in resin-buffered solutions and in soil. Plant and Soil 306:69-84. Degryse F., Smolders E., Parker D.R. (2009a) Partitioning of metals (Cd, Co, Cu, Ni, Pb, Zn) in soils: concepts, methodologies, prediction and applications - a review. European Journal of Soil Science 60:590-612. Degryse F., Broos K., Smolders E., Merckx R. (2003) Soil solution concentration of Cd and Zn can be predicted with a CaCl2 soil extract. European Journal of Soil Science 54:149- 157. Degryse F., Vlassak V., Smolders E., Merckx R. (2007) Mobilization of Cd upon acidification of agricultural soils: column study and field modelling. European Journal of Soil Science 58:152-165. Degryse F., Smolders E., Zhang H., Davison W. (2009b) Predicting availability of mineral elements to plants with the DGT technique: a review of experimental data and interpretation by modelling. Environmental Chemistry 6:198-218.
167 Deram A., Denayer F.O., Petit D., Van Haluwyn C. (2006) Seasonal variations of cadmium and zinc in Arrhenatherum elatius, a perennial grass species from highly contaminated soils. Environmental Pollution 140:62-70. Dheri G.S., Brar M.S., Malhi S.S. (2007) Influence of phosphorus application on growth and cadmium uptake of spinach in two cadmium-contaminated soils. Journal of Plant Nutrition and Soil Science-Zeitschrift Fur Pflanzenernahrung Und Bodenkunde 170:495-499. Doig L.E., Liber K. (2007) Nickel speciation in the presence of different sources and fractions of dissolved organic matter. Ecotoxicology and Environmental Safety 66:169-177. Dumestre A., Sauve S., McBride M., Baveye P., Berthelin J. (1999) Copper speciation and microbial activity in long-term contaminated soils. Archives of Environmental Contamination and Toxicology 36:124-131. Dunbar K.R., McLaughlin M.J., Reid R.J. (2003) The uptake and partitioning of cadmium in two cultivars of potato (Solanum tuberosum L.). Journal of Experimental Botany 54:349-354. Dussault M., Becaert V., Francois M., Sauve S., Deschenes L. (2008) Effect of copper on soil functional stability measured by relative soil stability index (RSSI) based on two enzyme activities. Chemosphere 72:755-762. El-Aziz R., Angle J.S., Chaney R.L. (1991) Metal tolerance of Rhizobium meliloti isolated from heavy-metal contaminated soils. Soil Biology and Biochemistry 23:795-798. Ernstberger H., Davison W., Zhang H., Tye A., Young S. (2002) Measurement and dynamic modeling of trace metal mobilization in soils using DGT and DIFS. Environmental Science and Technology 36:349-354. Ernstberger H., Zhang H., Tye A., Young S., Davison W. (2005) Desorption kinetics of Cd, Zn, and Ni measured in soils by DGT. Environmental Science and Technology 39:1591-1597. Ettler V., Mihaljevic M., Sebek O., Grygar T. (2007) Assessment of single extractions for the determination of mobile forms of metals in highly polluted soils and sediments - Analytical and thermodynamic approaches. Analytica Chimica Acta 602:131-140. Fairbrother A., Wenstel R., Sappington K., Wood W. (2007) Framework for metals risk assessment. Ecotoxicology and Environmental Safety 68:145-227. Georgieva S.S., McGrath S.P., Hooper D.J., Chambers B.S. (2002) Nematode communities under stress: the long-term effects of heavy metals in soil treated with sewage sludge. Applied Soil Ecology 20:27-42. Giller K.E., Witter E., McGrath S.P. (2009) Heavy metals and soil microbes. Soil Biology and Biochemistry 41:2031-2037. Goss M.J., Ehlers W. (2009) The role of lysimeters in the development of our understanding of soil water and nutrient dynamics in ecosystems. Soil Use and Management 25:213- 223. Gray C.W., McLaren R.G. (2005) The effect of ryegrass variety on trace metal uptake. New Zealand Journal of Agricultural Research 48:285-292. Gray C.W., McLaren R.G. (2006) Soil factors affecting heavy metal solubility in some New Zealand soils. Water Air and Soil Pollution 175:3-14.
168 Gray C.W., McLaren R.G., Shiowatana J. (2003) The determination of labile cadmium in some biosolids-amended soils by isotope dilution plasma mass spectrometry. Australian Journal of Soil Research 41:589-597. Gray C.W., McLaren R.G., Roberts A.H.C., Condron L.M. (1999a) Solubility, sorption and desorption of native and added cadmium in relation to properties of soils in New Zealand. European Journal of Soil Science 50:127-137. Gray C.W., McLaren R.G., Roberts A.H.C., Condron L.M. (1999b) Cadmium phytoavailability in some New Zealand soils. Australian Journal of Soil Research 37:461-477. Gray C.W., McLaren R.G., Roberts A.H.C., Condron L.M. (1999c) The effect of long-term phosphatic fertiliser applications on the amounts and forms of cadmium in soils under pasture in New Zealand. Nutrient Cycling in Agroecosystems 54:267-277. Gray C.W., Moot D.J., McLaren R.G., Reddecliffe T. (2002) Effect of nitrogen fertiliser applications on cadmium concentrations in durum wheat (Triticum turgidum) grain. New Zealand Journal of Crop and Horticultural Science 30:291-299. Gray C.W., McLaren R.G., Gunther D., Sinaj S. (2004) An assessment of cadmium availability in cadmium-contaminated soils using isotope exchange kinetics. Soil Science Society of America Journal 68:1210-1217. Gray C.W., McLaren, R.G., Roberts, A.H.C., Condron, L.M. (1999d) Effect of soil pH on cadmium phytoavailability in some New Zealand soils. New Zealand Journal of Crop and Horticultural Science 27:169-179. Gregory P.J. (2006) Roots, rhizosphere and soil: the route to a better understanding of soil science? European Journal of Soil Science 57:2-12. Griffiths E. (1974) Soils of part of the Port Hills and adjacent plains, Canterbury, New Zealand., Soil Bureau Bulletin 35., Department of Scientific and Industrial Research, New Zealand, Havelock North. pp. 36. Gutierrez-Ruiz M., Romero F.M., Gonzalez-Hernandez G. (2007) Soils and sediments affected by the dispersion of metal sulfide from inactive tailings in the Santa Barbara mining area, Chihuahua, Mexico. Revista Mexicana De Ciencias Geologicas 24:170- 184. Hamon R., Wundke J., McLaughlin M., Naidu R. (1997) Availability of zinc and cadmium to different plant species. Australian Journal of Soil Research 35:1267-1277. Harper M.P., Davison W., Tych W. (2000) DIFS - a modelling and simulation tool for DGT induced trace metal remobilisation in sediments and soils. Environmental Modelling and Software 15:55-66. Heenan D.P., Campbell L.C. (1981) Influence of potassium and manganese on growth and uptake of magnesium by soybeans (glycine-max (l) merr cv bragg). Plant and Soil 61:447-456. Helyar K.R., Porter W.M. (1989) Soil acidification, its measurement and the processes involved. Soil acidity and plant growth:61-101. Himmelbauer M.L., Puschenreiter M., Schnepf A., Loiskandl W., Wenzel W.W. (2005) Root morphology of Thlaspi goesingense Halacsy grown on a serpentine Soil. Journal of
169 Plant Nutrition and Soil Science-Zeitschrift Fur Pflanzenernahrung Und Bodenkunde 168:138-144. Hirsch P.R., Jones M.J., McGrath S.P., Giller K.E. (1993) Heavy metals from past applications of sewage sludge decrease the genetic diversity of rhizobium leguminosarum biovar trifolii populations. Soil Biology and Biochemistry 25:1485- 1490. Homer F.A., Reeves R.D., Brooks R.R., Baker A.J.M. (1991) Characterization of the nickel- rich extract from the nickel hyperaccumulator dichapetalum-gelonioides. Phytochemistry 30:2141-2145. Hooda P.S. (2007) Speciation and bioavailability of metal contaminants in soils: Concepts, approaches and challenges. Progress in Environmental Science and Technology, Vol I:1165-1168. Hooda P.S., Alloway B.J. (1994a) Sorption of Cd and Pb by Selected Temperate and Semiarid Soils - Effects of Sludge Application and Aging of Sludged Soils. Water Air and Soil Pollution 74:235-250. Hooda P.S., Alloway B.J. (1994b) The plant availability and DTPA extractability of trace metals in sludge-amended soils. The Science of the Total Environment 149:39-51. Hooda P.S., Zhang H., Davison W., Edwards A.C. (1999) Measuring bioavailable trace metals by diffusive gradients in thin films (DGT): soil moisture effects on its performance in soils. European Journal of Soil Science 50:285-294. Horswell J., Speir T. (2006) Arsenic phytotoxicity: effect on crop yield and crop quality. Managing arsenic in the environment: from soil to human health:183-207. Horswell J., Speir T.W., van Schaik A.P. (2003) Bio-indicators to assess impacts of heavy metals in land-applied sewage sludge. Soil Biology and Biochemistry 35:1501-1505. Horswell J., Weitz H.J., Percival H.J., Speir T.W. (2006) Impact of heavy metal amended sewage sludge on forest soils as assessed by bacterial and fungal biosensors. Biology and Fertility of Soils 42:569-576. Horswell J., Ambrose V., Clucas L., Leckie A., Clinton P., Speir T.W. (2007) Survival of Escherichia coli and Salmonella spp. after application of sewage sludge to a Pinus radiata forest. Journal of Applied Microbiology 103:1321-1331. Houba V.J.G., Temminghoff E.J.M., Gaikhorst G.A., van Vark W. (2000) Soil analysis procedures using 0.01 M calcium chloride as extraction reagent. Communications in Soil Science and Plant Analysis 31:1299-1396. Hough R.L., Young S.D., Crout N.M.J. (2003) Modelling of Cd, Cu, Ni, Pb and Zn uptake, by winter wheat and forage maize, from a sewage disposal farm. Soil Use and Management 19:19-27. Hough R.L., Tye A.M., Crout N.M.J., McGrath S.P., Zhang H., Young S.D. (2005) Evaluating a 'Free Ion Activity Model' applied to metal uptake by Lolium perenne L. grown in contaminated soils. Plant and Soil 270:1-12. Idris R., Trifonova R., Puschenreiter M., Wenzel W.W., Sessitsch A. (2004) Bacterial communities associated with flowering plants of the Ni hyperaccumulator Thlaspi goesingense. Applied and Environmental Microbiology 70:2667-2677.
170 Idris R., Kuffner M., Bodrossy L., Puschenreiter M., Monchy S., Wenzel W.W., Sessitsch A. (2006) Characterization of Ni-tolerant methylobacteria associated with the hyperaccumulating plant Thlaspi goesingense and description of Methylobacterium goesingense sp nov. Systematic and Applied Microbiology 29:634-644. Imtiaz M., Alloway B.J., Aslam M., Memon M.Y., Khan P., Siddiqui S.U.H., Shah S.K.H. (2006) Zinc sorption in selected soils. Communications in Soil Science and Plant Analysis 37:1675-1688. Jones D.L., Hodge A., Kuzyakov Y. (2004) Plant and mycorrhizal regulation of rhizodeposition. New Phytologist 163:459-480. Kalis E.J.J., Temminghoff E.J.M., Weng L.P., Van Riemsdijk W.H. (2006) Effects of humic acid and competing cations on metal uptake by Lolium perenne. Environmental Toxicology and Chemistry 25:702-711. Khan M.A., Noora J., Qudsia S., Ahmad S.R., Anisur R. (2006) Effect of different organic materials and chemical fertilizers on the yield of wheat and physical properties of soil. Sarhad Journal of Agriculture 22:437-441. Khan M.A.R., Bolan N.S., Mackay A.D. (2005) Soil test to predict the copper availability in pasture soils. Communications in Soil Science and Plant Analysis 36:2601-2624. Kiekens L., Cottenie A. (1985) Principles of investigations on the mobility and plant uptake of heavy metals. Chemical methods for assessing bio-available metals in sludge and soils.:32-41. Kim B.J., McBride M.B., Richards B.K., Steenhuis T.S. (2007) The long-term effect of sludge application on Cu, Zn, and Mo behavior in soils and accumulation in soybean seeds. Plant and Soil 299:227-236. Knights J.S., Zhao F.J., McGrath S.P., Magan N. (2001) Long-term effects of land use and fertiliser treatments on sulphur transformations in soils from the Broadbalk experiment. Soil Biology and Biochemistry 33:1797-1804. Kopittke P.M., Dart P.J., Menzies N.W. (2007) Toxic effects of low concentrations of Cu on nodulation of cowpea (Vigna unguiculata). Environmental Pollution 145:309-315. Kovacs B., Prokisch, J., Gyori, Z., Kovacs, A.B., Palenscar, A.J. (2000) Studies on soil sample preparation for inductively coupled plasma atomic emission spectrometry analysis. Communications in Soil Science and Plant Analysis 31:1949-1963. Kovarikova V., Docekalova H., Docekal B., Podborska M. (2007) Use of the diffusive gradients in thin films technique (DGT) with various diffusive gels for characterization of sewage sludge-contaminated soils. Analytical and Bioanalytical Chemistry 389:2303- 2311. Kozlova T., Wood C.M., McGeer J.C. (2009) The effect of water chemistry on the acute toxicity of nickel to the cladoceran Daphnia pulex and the development of a biotic ligand model. Aquatic Toxicology 91:221-8. Krpata D., Fitz W., Peintner U., Langer I., Schweiger P. (2009) Bioconcentration of zinc and cadmium in ectomycorrhizal fungi and associated aspen trees as affected by level of pollution. Environmental Pollution 157:280-286.
171 Lair G.J., Gerzabek M.H., Haberhauer G. (2007) Retention of copper, cadmium and zinc in soil and its textural fractions influenced by long-term field management. European Journal of Soil Science 58:1145-1154. Lair G.J., Gerzabek M.H., Haberhauer G., Jakusch M., Kirchmann H. (2006) Response of the sorption behavior of Cu, Cd, and Zn to different soil management. Journal of Plant Nutrition and Soil Science-Zeitschrift Fur Pflanzenernahrung Und Bodenkunde 169:60- 68. Lehto N.J., Davison W., Zhang H., Tych W. (2006a) Theoretical comparison of how soil processes affect uptake of metals by diffusive gradients in thinfilms and plants. Journal of Environmental Quality 35:1903-1913. Lehto N.J., Davison W., Zhang H., Tych W. (2006b) Analysis of micro-nutrient behaviour in the rhizosphere using a DGT parameterised dynamic plant uptake model. Plant and Soil 282:227-238. Lehto N.J., Davison W., Zhang H., Tych W. (2006c) An evaluation of DGT performance using a dynamic numerical model. Environmental Science and Technology 40:6368- 6376. Lock K., Janssen C.R. (2001) Ecotoxicity of zinc in spiked artificial soils versus contaminated field soils. Environmental Science and Technology 35:4295-4300. Lock K., Criel P., De Schamphelaere K.A.C., Van Eeckhout H., Janssen C.R. (2007) Influence of calcium, magnesium, sodium, potassium and pH on copper toxicity to barley (Hordeum vulgare). Ecotoxicology and Environmental Safety 68:299-304. Lofts S., Spurgeon D.J., Svendsen C., Tipping E. (2004) Deriving soil critical limits for Cu, Zn, Cd, and Ph: A method based on free ion concentrations. Environmental Science and Technology 38:3623-3631. Lofts S., Chapman P.M., Dwyer R., McLaughlin M.J., Schoeters I., Sheppard S.C., Adams W.J., Alloway B.J., Antunes P.M.C., Campbell P.G.C., Davies B., Degryse F., De Vries W., Farley K.J., Garrett R.G., Green A., Groenenberg B.J., Hale B., Harrass M., Hendershot W.H., Keller A., Lanno R., Liang T., Liu W.X., Ma Y.B., Menzie C., Moolenaar S.W., Piatkiewicz W., Reimann C., Rieuwerts J.S., Santore R.C., Sauve S., Schuetze G., Schlekat C., Skeaff J., Smolders E., Tao S., Wilkins J., Zhao F.J. (2007) Critical loads of metals and other trace elements to terrestrial environments. Environmental Science and Technology 41:6326-6331. Loganathan P., Hedley M.J., Grace N.D., Lee J., Cronin S.J., Bolan N.S., Zanders J.M. (2003) Fertiliser contaminants in New Zealand grazed pasture with special reference to cadmium and fluorine: a review. Australian Journal of Soil Research 41:501-532. Lombnaes P., Chang A.C., Singh B.R. (2008) Organic ligand, competing cation, and pH effects on dissolution of zinc in soils. Pedosphere 18:92-101. Longhurst R.D., Roberts A.H.C., Waller J.E. (2004) Concentrations of arsenic, cadmium, copper, lead, and zinc in New Zealand pastoral topsoils and herbage. New Zealand Journal of Agricultural Research 47:23-32. Lu H.L., Yan C.L., Liu J.C. (2007) Low-molecular-weight organic acids exuded by Mangrove (Kandelia candel (L.) Druce) roots and their effect on cadmium species change in the rhizosphere. Environmental and Experimental Botany 61:159-166.
172 Ma Y., Mani R., Freitas H. (2009) Isolation and characterization of Ni mobilizing PGPB from serpentine soils and their potential in promoting plant growth and Ni accumulation by Brassica spp. Chemosphere 75:719-725. Macnicol R.D., Beckett P.H.T. (1985) Critical Tissue Concentrations of Potentially Toxic Elements. Plant and Soil 85:107-129. Marschner B., Kalbitz K. (2003) Controls of bioavailability and biodegradability of dissolved organic matter in soils. Geoderma 113:211-235. Marschner H. (1995) Mineral nutrient acquisition in nonmycorrhizal and mycorrhizal plants. Proceedings of the international colloquium on Bioindication of forest site pollution: development of methodology and training (BIOFOSP), 22-31 August 1995, Ljubljana, Slovenia.:101-106. Marschner H., Marschner H. (1995) Mineral nutrition of higher plants, Second edition. Mineral nutrition of higher plants, Second edition:xv+889p. McBride M., Sauve S., Hendershot W. (1997a) Solubility control of Cu, Zn, Cd and Pb in contaminated soils. European Journal of Soil Science 48:337-346. McBride M.B. (1994) Environmental chemistry of soils. Environmental chemistry of soils.:vii + 406 pp. McBride M.B. (1995) Toxic Metal Accumulation from Agricultural Use of Sludge - Are Usepa Regulations Protective. Journal of Environmental Quality 24:5-18. McBride M.B., Pitiranggon M., Kim B. (2009) A Comparison of Tests for Extractable Copper and Zinc in Metal-Spiked and Field-Contaminated Soil. Soil Science 174:439-444. McBride M.B., Nibarger E.A., Richards B.K., Steenhuis T. (2003) Trace metal accumulation by red clover grown on sewage sludge-amended soils and correlation to Mehlich 3 and calcium chloride-extractable metals. Soil Science 168:29-38. McBride M.B., Richards B.K., Steenhuis T., Russo J.J., Sauve S. (1997b) Mobility and solubility of toxic metals and nutrients in soil fifteen years after sludge application. Soil Science 162:487-500. McIlveen W.D., Negusanti J.J. (1994) Nickel in the terrestrial environment. Science of the Total Environment 148:109-138. McKeague J.A., Day, J.H. (1966) Dithionite and oxalate-extractable Fe and Al as aids in differentiating various classes of soils. . Canadian Journal of Soil Science 46:13-22. McLaren R. (2005) Effects of annual applications of biosolids materials on soils, plants, and drainage leachates - a lysimeter study results from the first two years, Sewage Biosolids - Safe, Beneficial and Acceptable Use of Land: Contract No. CO3X0302, Institue of Environmental Science and Research Ltd. pp. 65. McLaren R. (2007) Fate of metals and nutrients added to a pasture soil in metal-spiked biosolids. Sewage Biosolids - Safe, Beneficial and Acceptable Use of Land: Contract No. CO3X0302, Institue of Environmental Science and Research Ltd. pp. 91. McLaren R.G., Crawford D.V. (1973a) Studies on soil copper .1. fractionation of copper in soils. Journal of Soil Science 24:172-181. McLaren R.G., Crawford D.V. (1973b) Studies on soil copper .2. specific adsorption of copper by soils. Journal of Soil Science 24:443-452.
173 McLaren R.G., Clucas L.M. (2001) Fractionation of copper, nickel, and zinc in metal-spiked sewage sludge. Journal of Environmental Quality 30:1968-1975. McLaren R.G., Clucas L.M., Speir T.W., van Schaik A.P. (2007) Distribution and movement of nutrients and metals in a Pinus radiata forest soil following applications of biosolids. Environmental Pollution 147:32-40. McLaren R.G., Clucas, L.M., Taylor, M.D., Hendry, T. (2003) Leaching of macronutrients and metals from undisturbed soils treated with metal-spiked sewage sludge 1. Leaching of macronutrients. Australian Journal of Soil Research 41:571-588. McLaren R.G., Clucas, L.M., Taylor, M.D., Hendry, T. (2004) Leaching of macronutrients and metals from undisturbed soils treated with metal-spiked sewage sludge 2. Leaching of metals. Australian Journal of Soil Research 42:459-471. McLaren R.G., Clucas, L.M., Taylor, M.D., Hendry, T. (2005) Leaching of macronutrients and metals from undisturbed soils treated with metal-spiked sewage sludge 3. Distribution of residual metals. Australian Journal of Soil Research 43:159-170. McLaren R.G., Swift, R.S., Quinn, B.F. (1984) EDTA-extractable copper, zinc and manganese in the soils of the Canterbury Plains. New Zealand Journal of Agricultural Research 27:207-217. McLaughlin M.J. (2002) Long term changes in cadmium bioavailability in soil. International Atomic Energy Agency Technical Documents (IAEA-TECDOCs):354-362. McLaughlin M.J., Tiller K.G., Smart M.K. (1997) Speciation of cadmium in soil solutions of saline/sodic soils and relationship with cadmium concentrations in potato tubers (Solanum tuberosum L). Australian Journal of Soil Research 35:183-198. McLaughlin M.J., Tiller K.G., Naidu R., Stevens D.P. (1996) Review: The behaviour and environmental impact of contaminants in fertilizers. Australian Journal of Soil Research 34:1-54. McLaughlin M.J., Zarcinas B.A., Stevens D.P., Cook N. (1999) Soil testing for heavy metals, International Symposium on Soil and Plant Analysis, Marcel Dekker Inc, Brisbane, Australia. pp. 1661-1700. McLaughlin M.J., Palmer L.T., Tiller K.G., Beech T.A., Smart M.K. (1994) Increased Soil- Salinity Causes Elevated Cadmium Concentrations in Field-Grown Potato-Tubers. Journal of Environmental Quality 23:1013-1018. McLaughlin M.J., Hamon R.E., McLaren R.G., Speir T.W., Rogers S.L. (2000) Review: A bioavailability-based rationale for controlling metal and metalloid contamination of agricultural land in Australia and New Zealand. Australian Journal of Soil Research 38:1037-1086. McLaughlin M.J., Maier N.A., Freeman K., Tiller K.G., Williams C.M.J., Smart M.K. (1995) Effect of Potassic and Phosphatic Fertilizer Type, Fertilizer Cd Concentration and Zinc Rate on Cadmium Uptake by Potatoes. Fertilizer Research 40:63-70. McLaughlin M.J., Whatmuff M., Warne M., Heemsbergen D., Barry G., Bell M., Nash D., Pritchard D. (2006) A field investigation of solubility and food chain accumulation of biosolid-cadmium across diverse soil types. Environmental Chemistry 3:428-432. McLaughlin M.J., Warne M.S.J., Stevens D.P., Whatmuff M.S., Heemsbergen D., Broos K., Barry G., Bell M.J., Nash D., Pritchard D., Penney N. (2007) Australia's National
174 Biosolid Research Program - how it came about, and what has it discovered? Water Practice & Technology 2:art 88. McNear D.H., Chaney R.L., Sparks D.L. (2007) The effects of soil type and chemical treatment on nickel speciation in refinery enriched soils: A multi-technique investigation. Geochimica et Cosmochimica Acta 71:2190-2208. Meers E., Du Laing G., Tack F.M.G., Verloo M.G. (2009) Heavy Metal Displacement by Exchangeable Bases (Ca, Mg, K, Na) in Soils and Sediments. Soil Science 174:202- 209. Meers E., Samson R., Tack F.M.G., Ruttens A., Vandegehuchte M., Vangronsveld J., Verloo M.G. (2007) Phytoavailability assessment of heavy metals in soils by single extractions and accumulation by Phaseolus vulgaris. Environmental and Experimental Botany 60:385-396. Meers E., Tack F.M.G., Van Slycken S., Ruttens A., Laing G.D., Vangronsveld J., Verloo M.G. (2008) Chemically assisted phytoextraction: A review of potential soil amendments for increasing plant uptake of heavy metals. International Journal of Phytoremediation 10:390-414. Meers E., Unamuno V., Vandegehuchte M., Vanbroekhoven K., Geebelen W., Samson R., Vangronsveld J., Diels L., Ruttens A., Du Laing G., Tack F. (2005) Soil-solution speciation of Cd as affected by soil characteristics in unpolluted and polluted soils. Environmental Toxicology and Chemistry 24:499-509. Menzies N.W., Donn M.J., Kopittke P.M. (2007) Evaluation of extractants for estimation of the phytoavailable trace metals in soils. Environmental Pollution 145:121-130. Merrington G., Smernik R.J. (2004) Cadmium sorption in biosolids amended soils: results from a field trial. Science of the Total Environment 327:239-247. Merrington G., Winder L., Green I. (1997) The bioavailability of Cd and Zn from soils amended with sewage sludge to winter wheat and subsequently to the grain aphid Sitobion avenae. Science of the Total Environment 205:245-254. Merrington G., Oliver I., Smernik R.J., McLaughlin M.J. (2003) The influence of sewage sludge properties on sludge-borne metal availability. Advances in Environmental Research 8:21-36. Molas J., Baran S. (2004) Relationship between the chemical form of nickel applied to the soil and its uptake and toxicity to barley plants (Hordeum vulgare L.). Geoderma 122:247-255. Munns R., Tester M. (2008) Mechanisms of salinity tolerance. Annual Review of Plant Biology 59:651-681. Naidu R., Rogers S., Gupta V.S.R., Kookana R.S., Bolan N.S., Adriano D.C. (2003) Bioavailability of metals in the soil plant environment and its potential role in risk assessment. Bioavailability, Toxicity and Risk Relationships in Ecosystems:21-57. Navas A., Machin J., Navas B. (1999) Use of biosolids to restore the natural vegetation cover on degraded soils in the badlands of Zaragoza (NE Spain). Bioresource Technology 69:199-205.
175 Nicholson F.A., Smith S.R., Alloway B.J., Carlton-Smith C., Chambers B.J. (2006) Quantifying heavy metal inputs to agricultural soils in England and Wales. Water and Environment Journal 20:87-95. Nolan A.L., McLaughlin M.J., Mason S.D. (2003) Chemical speciation of Zn, Cd, Cu, and Pb in pore waters of agricultural and contaminated soils using Donnan dialysis. Environmental Science & Technology 37:90-98. Nolan A.L., Zhang H., McLaughlin M.J. (2005) Prediction of zinc, cadmium, lead, and copper availability to wheat in contaminated soils using chemical speciation, diffusive gradients in thin films, extraction, and isotopic dilution techniques. Journal of Environmental Quality 34:496-507. Nolan A.L., Ma Y., Lombi E., McLaughlin M.J. (2009) Speciation and isotopic exchangeability of nickel in soil solution. Journal of Environmental Quality 38:485-492. Nwachukwu O.I., Pulford I.D. (2008) Comparative effectiveness of selected adsorbant materials as potential amendments for the remediation of lead-, copper- and zinc- contaminated soil. Soil Use and Management 24:199-207. Nwachukwu O.I., Pulford I.D. (2009) Soil metal immobilization and ryegrass uptake of lead, copper and zinc as affected by application of organic materials as soil amendments in a short-term greenhouse trial. Soil Use and Management 25:159-167. NZWWA. (2003) Guidelines for the safe application of biosolids to land in New Zealand in: N. Z. W. W. A. (NZWWA) (Ed.), New Zealand Waste Water Association (NZWWA), Wellington, New Zealand, Wellington. Oliver I.W., Merrington G., McLaughlin M.J. (2004) Australian biosolids: characterization and determination of available copper. Environmental Chemistry 1:116-124. Oliver I.W., McLaughlin M.J., Merrington G. (2005) Temporal trends of total and potentially available element concentrations in sewage biosolids: a comparison of biosolid surveys conducted 18 years apart. Science of the Total Environment 337:139-145. Öncel I., Keles Y., Üstün A.S. (2000) Interactive effects of temperature and heavy metal stress on the growth and some biochemical compounds in wheat seedlings. Environmental Pollution 107:315-320. Oorts K., Smolders E., Degryse F., Buekers J., Gasco G., Cornelis G., Mertens J. (2008) Solubility and toxicity of antimony trioxide (Sb2O3) in soil. Environmental Science and Technology 42:4378-4383. Oporto C., Smolders E., Degryse F., Verheyen L., Vandecasteele C. (2009) DGT-measured fluxes explain the chloride-enhanced cadmium uptake by plants at low but not at high Cd supply. Plant and Soil 318:127-135. Ozkutlu F., Ozturk L., Erdem H., McLaughlin M., Cakmak I. (2007) Leaf-applied sodium chloride promotes cadmium accumulation in durum wheat grain. Plant and Soil 290:323-331. Palmer C.M., Guerinot M.L. (2009) Facing the challenges of Cu, Fe and Zn homeostasis in plants. Nature Chemical Biology 5:333-340. Ponizovsky A.A., Thakali S., Allen H.E., Di Toro D.M., Ackerman A.J., Metzler D.M. (2008) Nickel partitioning in acid soils at low moisture content. Geoderma 145:69-76.
176 Puschenreiter M., Schnepf A., Millan I.M., Fitz W.J., Horak O., Klepp J., Schrefl T., Lombi E., Wenzel W.W. (2005) Changes of Ni biogeochemistry in the rhizosphere of the hyperaccumulator Thlaspi goesingense. Plant and Soil 271:205-218. Rachou J., Hendershot W., Sauve S. (2004) Effects of pH on fluxes of cadmium in soils measured by using diffusive gradients in thin films. Communications in Soil Science and Plant Analysis 35:2655-2673. Reichman S.M. (2002) The Responses of Plants to Metal Toxicity: A review focusing on Copper, Manganese and Zinc, Australian Minerals and Energy Environment Foundation Melbourne, Australia. pp. 54. Reichman S.M., Menzies N.W., Asher C.J., Mulligan D.R. (2006) Responses of four Australian tree species to toxic concentrations of copper in solution culture. Journal of Plant Nutrition 29:1127-1141. Reid R., Hayes J., Kwang W.J. (2003a) Mechanisms and Control of Nutrient Uptake in Plants, International Review of Cytology, Academic Press. pp. 73-114. Reid R.J. (2001) Mechanisms of micronutrient uptake in plants. Australian Journal of Plant Physiology 28:659-666. Reid R.J., Dunbar K.R., McLaughlin M.J. (2003b) Cadmium loading into potato tubers: the roles of the periderm, xylem and phloem. Plant Cell and Environment 26:201-206. Roberts A.H.C., Longhurst R.D. (2002) Cadmium cycling in sheep-grazed hill-country pastures. New Zealand Journal of Agricultural Research 45:103-112. Roberts A.H.C., Longhurst R.D., Brown M.W. (1994) Cadmium status of soils, plants, and grazing animals in New Zealand. New Zealand Journal of Agricultural Research 37:119-129. Rooney C.P., Zhao F.J., McGrath S.P. (2006) Soil factors controlling the expression of copper toxicity to plants in a wide range of European soils. Environmental Toxicology and Chemistry 25:726-732. Rooney C.P., McLaren R.G., Condron L.M. (2007a) Control of lead solubility in soil contaminated with lead shot: Effect of soil pH. Environmental Pollution 149:149-157. Rooney C.P., Zhao F.J., McGrath S.P. (2007b) Phytotoxicity of nickel in a range of European soils: Influence of soil properties, Ni solubility and speciation. Environmental Pollution 145:596-605. Ruttens A., Mench M., Colpaert J.V., Boisson J., Carleer R., Vangronsveld J. (2006) Phytostabilization of a metal contaminated sandy soil. I: Influence of compost and/or inorganic metal immobilizing soil amendments on phytotoxicity and plant availability of metals. Environmental Pollution 144:524-532. Santibanez C., Verdugo C., Ginocchio R. (2008) Phytostabilization of copper mine tailings with biosolids: Implications for metal uptake and productivity of Lolium perenne. Science of the Total Environment 395:1-10. Sarathy V., Allen H.E. (2005) Copper complexation by dissolved organic matter from surface water and wastewater effluent. Ecotoxicology and Environmental Safety 61:337-344. Sauve S., McBride M.B., Hendershot W.H. (1995) Ion-selective electrode measurements of copper (II) activity in contaminated soils. Archives of Environmental Contamination and Toxicology 29:373-379.
177 Sauve S. (2001) Speciation of metals in soils. Bioavailability of Metals in Terrestrial Ecosystems: Importance of Partitioning for Bioavailability to Invertebrates, Microbes, and Plants:7-38. Sauve S. (2003) The role of chemical speciation in bioavailability. Bioavailability, Toxicity and Risk Relationships in Ecosystems:59-82. Sauve S., Parker D.R. (2005) Chemical speciation of trace elements in soil solution. Chemical processes in soils:655-688. Sauve S., Hendershot W., Allen H.E. (2000a) Solid-solution partitioning of metals in contaminated soils: Dependence on pH, total metal burden, and organic matter. Environmental Science and Technology 34:1125-1131. Sauve S., McBride M.B., Norvell W.A., Hendershot W.H. (1997) Copper solubility and speciation of in situ contaminated soils: Effects of copper level, pH and organic matter. Water Air and Soil Pollution 100:133-149. Sauve S., Norvell W.A., McBride M., Hendershot W. (2000b) Speciation and complexation of cadmium in extracted soil solutions. Environmental Science and Technology 34:291- 296. Shi Z.Q., Allen H.E., Di Toro D.M., Lee S.Z., Meza D.M.F., Lofts S. (2007) Predicting cadmium adsorption on soils using WHAM VI. Chemosphere 69:605-612. Slaveykova V.I., Dedieu K., Parthasarathy N., Hajdu R. (2009) Effect of Competing Ions and Complexing Organic Substances on the Cadmium Uptake by the Soil Bacterium Sinorhizobium Meliloti. Environmental Toxicology and Chemistry 28:741-748. Sloan J.J., Dowdy R.H., Dolan M.S., Linden D.R. (1997) Long-term effects of biosolids applications on heavy metal bioavailability in agricultural soils. Journal of Environmental Quality 26:966-974. Smart M.C., G. Zarcinas, B. Stevens, D. Barry, G. Cockley, T. McLaughlin, M. . (2004) CSIRO Land and Water's Methods Manual for ACIAR Project No. LWR1/1998/119, in: D. Stevens (Ed.), Impact of heavy metals on sustainability of fertilization and waste recycling in peri-urban and intensive agriculture in south-east Asia, CSIRO Land and Water. Smith S.R. (2009) A critical review of the bioavailability and impacts of heavy metals in municipal solid waste composts compared to sewage sludge. Environment International 35:142-156. Smolders E., Oorts K., van Sprang P., Schoeters I., Janssen C.R., McGrath S.P., McLaughlin M.J. (2009) Toxicity of trace metals in soil as affected by soil type and aging after contamination: using calibrated bioavailability models to set ecological soil standards. Environmental Toxicology and Chemistry 28:1633-1642. Sochaczewski L., Tych W., Davison B., Zhang H. (2007) 2D DGT induced fluxes in sediments and soils (2D DIFS). Environmental Modelling and Software 22:14-23. Speir T.W., van Schaik A.P., Lloyd-Jones A.R., Kettles H.A. (2003a) Temporal response of soil biochemical properties in a pastoral soil after cultivation following high application rates of undigested sewage sludge. Biology and Fertility of Soils 38:377-385.
178 Speir T.W., Van Schaik A.P., Percival H.J., Close M.E., Pang L.P. (2003b) Heavy metals in soil, plants and groundwater following high-rate sewage sludge application to land. Water Air and Soil Pollution 150:319-358. Speir T.W., Horswell J., van Schaik A.P., McLaren R.G., Fietje G. (2004) Composted biosolids enhance fertility of a sandy loam soil under dairy pasture. Biology and Fertility of Soils 40:349-358. Speir T.W., van Schaik A.P., Hunter L.C., Ryburn J.L., Percival H.J. (2007) Attempts to derive EC50 values for heavy metals from land-applied Cu-Ni-, and Zn-spiked sewage sludge. Soil Biology and Biochemistry 39:539-549. Stacey S.P., McLaughlin M.J., Cakmak I., Hetitiarachchi G.M., Scheckel K.G., Karkkainen M. (2008) Root uptake of lipophilic zinc-rhamnolipid complexes. Journal of Agricultural and Food Chemistry 56:2112-2117. Stephan C.H., Courchesne F., Hendershot W.H., McGrath S.P., Chaudri A.M., Sappin-Didier V., Sauve S. (2008) Speciation of zinc in contaminated soils. Environmental Pollution 155:208-216. Stevenson B.G., Parkinson C.M., Mitchell M.J. (1984) Effect of Sewage-Sludge on Decomposition Processes in Soils. Pedobiologia 26:95-105. Stevenson F.J. (1994) Humus chemistry: genesis, composition, reactions. Humus chemistry: genesis, composition, reactions.:xiii + 496 pp. Stuczynski T., Siebielec G., Daniels W.L., McCarty G., Chaney R.L. (2007) Biological aspects of metal waste reclamation with biosolids. Journal of Environmental Quality 36:1154-1162. Sullivan T.S., Stromberger M.E., Paschke M.W. (2006a) Parallel shifts in plant and soil microbial communities in response to biosolids in a semi-arid grassland. Soil Biology and Biochemistry 38:449-459. Sullivan T.S., Stromberger M.E., Paschke M.W., Ippolito J.A. (2006b) Long-term impacts of infrequent biosolids applications on chemical and microbial properties of a semi-arid rangeland soil. Biology and Fertility of Soils 42:258-266. Sun Q., Wang X.R., Ding S.M., Yuan X.F. (2005) Effects of exogenous organic chelators on phytochelatins production and its relationship with cadmium toxicity in wheat (Triticum aestivum L.) under cadmium stress. Chemosphere 60:22-31. Taiz L., Zeiger E., Taiz L., Zeiger E. (1998) Plant physiology, Second edition. Plant physiology, Second edition:xxvi+792p. Tambasco G., Sauve S., Cook N., McBride M., Hendershot W. (2000) Phytoavailability of Cu and Zn to lettuce (Lactuca sativa) in contaminated urban soils. Canadian Journal of Soil Science 80:309-317. Tiller KG., Honeysett JL., Devries MPC. (1972) Soil Zn and its uptake by plants. II. Soil chemistry in relation tp prediction of availability. Australian Journal of Soil Research 10:165-182. Tipping E. (1994) WHAM: a chemical equilibrium model and computer code for waters, sediments, and soils incorporating a discrete site/electrostatic model of ion-binding by humic substances. Computers and Geoscience 20:973-1023.
179 Tipping E. (1998) Humic ion-binding Model VI: an improved description of the interactions of protons and metal ions with humic substances. Aquatic Geochemistry 4:3-48. Tipping E. (2005) Modelling Al competition for heavy metal binding by dissolved organic matter in soil and surface waters of acid and neutral pH. Geoderma 127:293-304. Tipping E., Rey-Castro C., Bryan S.E., Hamilton-Taylor J. (2002) Al(III) and Fe(III) binding by humic substances in freshwaters, and implications for trace metal speciation. Geochimica et Cosmochimica Acta 66:3211-3224. Tipping E., Smith E.J., Lawlor A.J., Hughes S., Stevens P.A. (2003a) Predicting the release of metals from ombrotrophic peat due to drought-induced acidification. Environmental Pollution 123:239-253. Tipping E., Rieuwerts J., Pan G., Ashmore M.R., Lofts S., Hill M.T.R., Farago M.E., Thornton I. (2003b) The solid-solution partitioning of heavy metals (Cu, Zn, Cd, Pb) in upland soils of England and Wales. Environmental Pollution 125:213-225. Tye A.M., Young S., Crout N.M.J., Zhang H., Preston S., Zhao F.J., McGrath S.P. (2004) Speciation and solubility of Cu, Ni and Pb in contaminated soils. European Journal of Soil Science 55:579-590. van Hees P.A.W., Jones D.L., Jentschke G., Godbold D.L. (2004) Mobilization of aluminium, iron and silicon by Picea abies and ectomycorrhizas in a forest soil. European Journal of Soil Science 55:101-111. van Herwijnen R., Laverye T., Poole J., Hodson M.E., Hutchings T.R. (2007a) The effect of organic materials on the mobility and toxicity of metals in contaminated soils. Applied Geochemistry 22:2422-2434. van Herwijnen R., Hutchings T.R., Ai-Tabbaa A., Moffat A.J., Johns M.L., Ouki S.K. (2007b) Remediation of metal contaminated soil with mineral-amended composts. Environmental Pollution 150:347-354. Vert G., Grotz N., Dedaldechamp F., Gaymard F., Guerinot M.L., Briata J.F., Curie C. (2002) IRT1, an Arabidopsis transporter essential for iron uptake from the soil and for plant growth. Plant Cell 14:1223-1233. Vogeler I., Vachey A., Deurer M., Bolan N. (2008) Impact of plants on the microbial activity in soils with high and low levels of copper. European Journal of Soil Biology 44:92- 100. Walter I., Martinez F., Alonso L., de Gracia J., Cuevas G. (2002) Extractable soil heavy metals following the cessation of biosolids application to agricultural soil. Environmental Pollution 117:315-321. Wang S.L., Nan Z.R., Liu X.W., Li Y., Qin S., Ding H.X. (2009) Accumulation and bioavailability of copper and nickel in wheat plants grown in contaminated soils from the oasis, northwest China. Geoderma 152:290-295. Warne M.S.J., Heemsbergen D., McLaughlin M., Bell M., Broos K., Whatmuff M., Barry G., Nash D., Pritchard D., Penney N. (2008a) Models for the field-based toxicity of copper and zinc salts to wheat in 11 Australian soils and comparison to laboratory-based models. Environmental Pollution 156:707-714. Warne M.S.J., Heemsbergen D., Stevens D., McLaughlin M., Cozens G., Whatmuff M., Broos K., Barry G., Bell M., Nash D., Pritchard D., Penney N. (2008b) Modelling the
180 toxicity of copper and zinc salts to wheat in 14 soils. Environmental Toxicology and Chemistry 27:786-792. Webb M. (1975) Cadmium. British Medical Bulletin 31:246-250. Webb M., Daniel M. (1975) Induced synthesis of metallothionein by pig kidney-cells invitro in response to cadmium. Chemico-Biological Interactions 10:269-276. Weggler-Beaton K., McLaughlin M.J., Graham R.D. (2000) Salinity increases cadmium uptake by wheat and Swiss chard from soil amended with biosolids. Australian Journal of Soil Research 38:37-45. Weggler K., McLaughlin M.J., Graham R.D. (2004) Effect of chloride in soil solution on the plant availability of biosolid-borne cadmium. Journal of Environmental Quality 33:496- 504. Welch R.M. (1995) Micronutrient nutrition of plants. Critical Reviews in Plant Sciences 14:49-82. Weng L.P., Lexmond T.M., Wolthoorn A., Temminghoff E.J.M., Van Riemsdijk W.H. (2003) Phytotoxicity and bioavailability of nickel: Chemical speciation and bioaccumulation. Environmental Toxicology and Chemistry 22:2180-2187. Wenzel W.W., Bunkowski M., Puschenreiter M., Horak O. (2003) Rhizosphere characteristics of indigenously growing nickel hyperaccumulator and excluder plants on serpentine soil. Environmental Pollution 123:131-138. Whitehead D.C. (2000) Nutrient elements in soils. Nutrient elements in grassland: soil-plant- animal relationships:15-40. Yeates G.W., Speir T.W., Taylor M.D., Clucas L., Van Schaik A.P. (2006) Nematode responses to biosolids incorporation in five soil types. Biology and Fertility of Soils 42:550-555. Yuan G. (2009) Copper, zinc, and nickel in soil solution affected by biosolids amendment and soil management. Australian Journal of Soil Research 47:305-310. Zarcinas B.A., Ishak C.F., McLaughlin M.J., Cozens G. (2004) Heavy metals in soils and crops in southeast Asia. 1. Peninsular Malaysia. Environmental Geochemistry and Health 26:343-357. Zhang H., Davison W. (2001) In situ speciation measurements. Using diffusive gradients in thin films (DGT) to determine inorganically and organically complexed metals. Pure and Applied Chemistry 73:9-15. Zhang H., Young S.D. (2005) Characterizing the availability of metals in contaminated soils. II. The soil solution. Soil Use and Management 21:459-467. Zhang H., Davison W., Miller S., Tych W. (1995) In-Situ High-Resolution Measurements of Fluxes of Ni, Cu, Fe, and Mn and Concentrations of Zn and Cd in Porewaters by Dgt. Geochimica et Cosmochimica Acta 59:4181-4192. Zhang H., Davison W., Knight B., McGrath S. (1998) In situ measurements of solution concentrations and fluxes of trace metals in soils using DGT. Environmental Science and Technology 32:704-710.
181 Zhang H., Lombi E., Smolders E., McGrath S. (2004) Kinetics of Zn release in soils and prediction of Zn concentration in plants using diffusive gradients in thin. Environmental Science and Technology 38:3608-3613. Zhang H., Zhao F.J., Sun B., Davison W., McGrath S.P. (2001) A new method to measure effective soil solution concentration predicts copper availability to plants. Environmental Science and Technology 35:2602-2607. Zhao F.J., Hamon R.E., McLaughlin M.J. (2001) Root exudates of the hyperaccumulator Thlaspi caerulescens do not enhance metal mobilization. New Phytologist 151:613-620. Zhao F.J., Rooney C.P., Zhang H., McGrath S.P. (2006) Comparison of soil solution speciation and diffusive gradients in thin-films measurement as an indicator of copper bioavailability to plants. Environmental Toxicology and Chemistry 25:733-742.
182
Appendix A. Soil chemistry, metal extractability and speciation modelling results for Chapter 4
Table A1 Background concentrations (mg kg-1) of Cd, Cu, Ni and Zn in the three soils in the presence and absence of biosolids.
Soils treatment Total metal (mg kg -1) Cd Cu Ni Zn Halkett reps Control 1 0.369848 4.926937 7.502624 35.13134 2 0.404749 5.367459 7.626058 37.05242 3 0.400334 5.398377 7.348242 37.87965 Biosolids 1 0.51108 8.059946 7.945998 42.95433 2 0.436582 7.225169 8.194083 43.42184 3 1.041934 7.083655 7.860276 41.93603
Summit Control 1 1.235911 26.79143 13.08214 94.56916 2 1.049686 22.6644 13.03059 82.92493 3 1.026504 23.23634 13.76114 105.0761 Biosolids 1 1.087835 28.49446 14.25174 116.1114 2 1.04696 26.11836 13.87075 91.44682 3 1.021605 24.7723 13.73196 129.4379
Wakanui Control 1 0.757055 7.09145 5.759329 42.84147 2 0.741557 8.936173 6.182531 57.28573 3 0.690199 13.43214 7.812716 67.77449 Biosolids 1 1.138935 13.85516 6.984689 77.09915 2 0.732194 11.43786 6.879556 61.09529 3 1.37032 10.98879 6.87614 61.25232
183 Dissolved organic carbon Soil solution pH 8 Halkett soil Halkett soil 600 Two weeks Two weeks ) Six months
-1 Six months 7 500
400 6 pH 300 5
200
4
DOC concentration (mg L 100
0 3 C B BLM LM BMM MM BHM HM BCd Cd C B BLM LM BMM MM BHM HM BCd Cd
8 600 Summit soil Two weeks Summit soil Two weeks ) Six months Six months -1 7 500
400 6 pH 300 5 200
4
DOC concentration (mg DOC concentration L 100
0 3 C B BLM LM BMM MM BHM HM BCd Cd C B BLM LM BMM MM BHM HM BCd Cd 8 600 Wakanui soil Two weeks Wakanui soil Two weeks
) Six months Six months
-1 7 500
400 6 pH 300 5
200
4
DOC concentration (mg L (mg concentration DOC 100
0 3 C B BLM LM BMM MM BHM HM BCd Cd C B BLM LM BMM MM BHM HM BCd Cd
Treatment Treatment Figure A1 Comparison of dissolved organic carbon (DOC) and pH in soil solution extracted from treated soils at two weeks and six months incubation time. Error bars denote SEM.
184 Ca(NO3)2 extractable Cd EDTA extractable Cd 12 12 Halkett soil Halkett soil
) 10 10 -1 Two weeks Two weeks Six months Six months 8 Total Cd 8 Total Cd
6 6
4 4
2 2 Cd concentration (mg kg
0 0 CBBLMLMBMMMMBHMHMBCdCd C B BLM LM BMM MM BHM HM BCd Cd
12 12 Summit soil Summit soil
) 10 10 -1 Two weeks Two weeks Six months Six months 8 8 Total Cd Total Cd
6 6
4 4
2
Cd concentration (mg kg 2
0 0 CBBLMLMBMMMMBHMHMBCdCd C B BLM LM BMM MM BHM HM BCd Cd
12 12 Wakanui soil Wakanui soil
) 10 10 -1 Two weeks Two weeks Six months Six months 8 Total Cd 8 Total Cd
6 6
4 4
2
Cd concentration (mg kg 2
0 0 CBBLMLMBMMMMBHMHMBCdCd C B BLM LM BMM MM BHM HM BCd Cd
Treatment Treatment
Figure A2 Comparison of Ca(NO3)2, EDTA and total recoverable Cd in soils at two weeks and six month incubation times. Error bars denote SEM.
185 Ca(NO3)2 extractable Cu EDTA extractable Cu 500 Halkett soil Halkett soil
) 30
-1 Two weeks 400 Two weeks Six months Six months Total Cu 300 20
200
10 100 Cu concentration (mg kg
0 0 C B BLM LM BMM MM BHM HM BCd Cd C B BLM LM BMM MM BHM HM BCd Cd
500 Summit soil Summit soil
) 30
-1 400 Two weeks Two weeks Six months Six months Total Cu 300 20
200
10 100 Cu concentration (mg kg
0 0 C B BLM LM BMM MM BHM HM BCd Cd CBBLMLMBMMMMBHMHMBCdCd
500 Wakanui soil Wakanui soil 30 )
-1 Two weeks 400 Two weeks Six months Six months Total Cu 300 20
200
10 100 Cu concentration (mg kg
0 0 C B BLM LM BMM MM BHM HM BCd Cd CBBLMLMBMMMMBHMHMBCdCd
Treatment Treatment
Figure A3 Comparison of Ca(NO3)2, EDTA and total recoverable Cu in soils at two weeks and six month incubation times. Error bars denote SEM. Total Cu is only shown with the EDTA extractable results. Note the difference in y-axis for Ca(NO3)2 extractable results.
186 Ca(NO3)2 extractable Ni EDTA extractable Ni 120 120 Halkett soil Halkett soil
) 100 Two weeks 100 Two weeks -1 Six months Six months 80 Total Ni 80 Total Ni
60 60
40 40
Ni concentration (mg kg Ni concentration 20 20
0 0 C B BLM LM BMM MM BHM HM BCd Cd C B BLM LM BMM MM BHM HM BCd Cd
120 120 Summit soil Summit soil
) 100 100 Two weeks
-1 Two weeks Six months Six months 80 Total Ni 80 Total Ni
60 60
40 40
Ni concentration (mg kg 20 20
0 0 C B BLM LM BMM MM BHM HM BCd Cd C B BLM LM BMM MM BHM HM BCd Cd
120 120 Wakanui soil Wakanui soil
) 100 100 -1 Two weeks Two weeks Six months Six months 80 80 Total Ni Total Ni
60 60
40 40
20 20 Ni concentration (mg kg concentration Ni
0 0 C B BLM LM BMM MM BHM HM BCd Cd C B BLM LM BMM MM BHM HM BCd Cd
Treatment Treatment
Figure A4 Comparison of Ca(NO3)2, EDTA and total recoverable Ni in soils at two weeks and six month incubation times. Error bars denote SEM.
187 Ca(NO3)2 extractable Zn EDTA extractable Zn
500 500 Halkett soil Halkett soil )
-1 400 Two weeks 400 Two weeks Six months Six months Total Zn Total Zn 300 300
200 200
100 100 Zn concentration (mg kg (mg Zn concentration
0 0 C B BLM LM BMM MM BHM HM BCd Cd C B BLM LM BMM MM BHM HM BCd Cd
500 500 Summit soil Summit soil )
-1 400 Two weeks 400 Two weeks Six months Six months Total Zn Total Zn 300 300
200 200
100 100 Zn concentration (mg kg
0 0 C B BLM LM BMM MM BHM HM BCd Cd C B BLM LM BMM MM BHM HM BCd Cd
500 500 Wakanui soil Wakanui soil )
-1 400 Two weeks 400 Two weeks Six months Six months Total Zn Total Zn 300 300
200 200
100 100 Zn concentration (mg kg Zn concentration
0 0 C B BLM LM BMM MM BHM HM BCd Cd C B BLM LM BMM MM BHM HM BCd Cd
Treatment Treatment
Figure A5 Comparison of Ca(NO3)2, EDTA and total recoverable Zn in treated soils at two weeks and six month incubation times. Error bars denote SEM.
188 Total Cd in soil solution and solution pH
Two weeks Six months 2.0 8 2.0 8 Halkett soil Cd Halkett soil Cd ) pH pH -1 7 7 1.5 1.5
6 6 1.0 1.0 pH 5 5
0.5 0.5 4 4 Cd concentration (mg L concentration Cd
0.0 3 0.0 3 C B BLM LM BMM MM BHM HM BCd Cd C B BLM LM BMM MM BHM HM BCd Cd
0.8 8 0.8 8 Summit soil Cd Summit soil Cd ) pH pH -1 7 7 0.6 0.6
6 6 0.4 0.4 pH 5 5
0.2 0.2 4 4 Cd concentration (mg L concentration Cd
0.0 3 0.0 3 C B BLM LM BMM MM BHM HM BCd Cd C B BLM LM BMM MM BHM HM BCd Cd
0.8 8 0.8 8 Wakanui soil Wakanui soil Cd ) pH -1 7 7 0.6 0.6
6 6
0.4 0.4 pH 5 5
0.2 0.2 4 4
Cd concentration (mg L concentration Cd Cd pH 0.0 3 0.0 3 C B BLM LM BMM MM BHM HM BCd Cd C B BLM LM BMM MM BHM HM BCd Cd
Treatment Treatment Figure A6 Comparison of Cd concentrations and pH changes in soil solution extracted from treated soils at two weeks and six month incubation times. Note the variation of the y-axis for Halkett soil graphs. Error bars denote SEM.
189 Total Cu concentration in soil solution and solution pH
Two weeks Six months 10 8 10 8 Halkett soil Cu Halkett soil Cu
) pH pH
-1 8 7 8 7
6 6 6 6
pH 4 5 4 5
2 4 2 4 Cu concentrationCu (mg L
0 3 0 3 C B BLM LM BMM MM BHM HM BCd Cd C B BLM LM BMM MM BHM HM BCd Cd
8 8 1.4 Summit soil Cu 1.4 Summit soil Cu
) pH pH
-1 1.2 7 1.2 7
1.0 1.0 6 6 0.8 0.8 pH 0.6 5 0.6 5
0.4 0.4 4 4 Cu concentration (mg L Cu concentration 0.2 0.2
0.0 3 0.0 3 C B BLM LM BMM MM BHM HM BCd Cd C B BLM LM BMM MM BHM HM BCd Cd
8 8 1.4 Wakanui soil Cu 1.4 Wakanui soil Cu
) pH pH
-1 1.2 7 1.2 7
1.0 1.0 6 6 0.8 0.8 pH 0.6 5 0.6 5
0.4 0.4 4 4 Cu concentration (mgL 0.2 0.2
0.0 3 0.0 3 C B BLM LM BMM MM BHM HM BCd Cd C B BLM LM BMM MM BHM HM BCd Cd
Treatment Treatment Figure A7 Comparison of Cu concentrations and pH changes in soil solution extracted from treated soils at two weeks and six month incubation times. Note the variation of the y axis for the Halkett soil graphs. Error bars denote SEM.
190 Total Ni concentration in soil solution and solution pH
Two weeks Six months 50 8 50 8 Halkett soil Ni Halkett soil Ni pH
) pH
-1 40 7 40 7
30 6 30 6 pH
20 5 20 5
10 4 10 4 Ni concentration (mg L Ni concentration
0 3 0 3 C B BLM LM BMM MM BHM HM BCd Cd C B BLM LM BMM MM BHM HM BCd Cd
50 8 50 8 Summit soil Ni Summit soil Ni
) pH pH
-1 40 7 40 7
30 6 30 6
pH 20 5 20 5
10 4 10 4 Ni concentration (mg L (mg concentration Ni
0 3 0 3 C B BLM LM BMM MM BHM HM BCd Cd CBBLMLMBMMMMBHMHMBCdCd
50 8 50 8 Wakanui soil Ni Wakanui soil Ni pH pH 40 7 40 7
30 6 30 6 pH
20 5 20 5
Ni concentration (mg/L) concentration Ni 10 4 10 4
0 3 0 3 CBBLMLMBMMMMBHMHMBCdCd C B BLM LM BMM MM BHM HM BCd Cd Treatment Treatment Figure A8 Comparison of Ni concentrations and pH changes in soil solution extracted from treated soils at two weeks and six month incubation times. Error bars denote SEM.
191 Total Zn in soil solution and solution pH
Two weeks Six months 10 8 200 8 Halkett soil Zn Halkett soil Zn
) pH pH
-1 8 7 7 150
6 6 6
100 pH 4 5 5
50 2 4
Zn concentration L mg concentration Zn 4
0 3 0 3 C B BLM LM BMM MM BHM HM BCd Cd C B BLM LM BMM MM BHM HM BCd Cd
50 8 50 8 Summit soil Zn Summit soil Zn
) pH pH
-1 40 7 40 7
30 6 30 6
pH 20 5 20 5
10 4 10 4 Zn concentration (mg L (mg concentration Zn
0 3 0 3 C B BLM LM BMM MM BHM HM BCd Cd CBBLMLMBMMMMBHMHMBCdCd
50 8 50 8 Wakanui soil Zn Wakanui soil Zn
) pH pH
-1 40 7 40 7
30 6 30 6
pH 20 5 20 5
10 4 10 4 Zn concentration(mg L
0 3 0 3 C B BLM LM BMM MM BHM HM BCd Cd C B BLM LM BMM MM BHM HM BCd Cd
Treatment Treatment Figure A9 Comparison of Zn concentrations and pH changes in soil solution extracted from treated soils at two weeks and six month incubation times. Note the variations of both y-axes of the Halkett soil graphs. Error bars denote SEM.
192 (A) (B) 108 105 Cd Cu 107
106 104 )
) 5
10 -1 -1
4
10 (L kg (L kg (L D 3
D 10 K
K 103 10 10
2 Log
Log 10
102 101 y = 0.65 + -0.93x, r ² 0.95 (B + MS) 100 y = 2.15 + -0.57x, r ² 0.63 (B + MS) y = 0.70 + -0.90x, r ² 0.95 (MS) y = 2.18 + -0.51, r ² 0.55 (MS) 10-1 101 10-8 10-7 10-6 10-5 10-4 10-3 10-2 10-1 100 101 0.0001 0.001 0.01 0.1 1 10 100
-1 -1 Log10 soil solution Cd concentration (mg L ) Log10 soil solution Cu concentration (mg L ) (C) 10000 Ni Biosolids + metal salts (MS) Fitted line (B + MS) 1000 Metal salts (MS) Fitted line (MS) ) -1 100 (L kg (L D K
10 10 Log
1 y = 1.58 + -0.68x, r ² 0.90 (B + MS) y = 1.57 + -0.72x, r ² 0.93 (MS) 0.1 0.001 0.01 0.1 1 10 100 1000
-1 Log10 soil solution Ni concentration (mg L ) Figure A10 Partitioning co-efficient (KD) for biosolids and non-biosolids amended metal spiked soils (A) results for Cd, (B) results for Cu and, (C) results for Ni. All relationships were significant (P<0.0001), n = 71 (Biosolids + MS) and n = 72 (MS).
193 Halkett soil - two weeks Halkett soil - six months
Cd Cd
BCd BCd
HM HM
BHM BHM
MM MM
BMM BMM
Treatment Treatment LM LM
BLM BLM
B B
C C
0 204060801000 20406080100
Summit soil - two weeks Summit soil - six months
Cd Cd
BCd BCd
HM HM
BHM BHM
MM MM
BMM BMM
Treatment LM LM
BLM BLM
B B
C C
0 204060801000 20406080100
Wakanui soil - two weeks Wakanui soil - six months
Cd Cd BCd BCd HM HM BHM BHM MM MM BMM BMM LM Treatment Treatment LM BLM BLM B B C C
0 204060801000 20406080100 Mean % of Cd species distribution Mean % of Cd species distribution
FA bound Cd Cd2+ CdSO4 Remaining inorganic species Figure A11 Mean % of Cd species present in soil solution as colloidal fulvic acid bound (FA), simple 2+ + - + 2 ion (Cd ), CdSO4 and remaining inorganic species (CdOH , Cd(OH)2, CdCl , CdCl2, CdHCO3 , CdCO3 -,
Cd(CO3)2). NB: for the two week graphs Cd was below detection limits for untreated HK and SM, as well as the biosolids treatment for HK.
194 Halkett soil - two weeks Halkett soil - six months
Cd Cd BCd BCd HM HM BHM BHM MM MM BMM BMM
Treatment LM LM BLM BLM B B C C
0 204060801000 20406080100 Summit soil - two weeks Summit soil - six months
Cd Cd BCd BCd HM HM BHM BHM MM MM BMM BMM
Treatment LM LM BLM BLM B B C C
0 204060801000 20406080100 Wakanui soil - two weeks Wakanui soil - six months
Cd Cd BCd BCd HM HM BHM BHM MM MM BMM BMM
Treatment LM LM BLM BLM B B C C
0 20 40 60 80 100 020406080100 Mean % of Cd distribution in soil Mean % of Cd distribution in soil
Particulate HA Particulate FA Fe-oxides Al-oxides Clay Aqueous phase
Figure A12 Mean percentage of Cd distribution on organic matter (particulate HA and FA), Fe and Al oxides, clay, and present in the aqueous phase for HK, SM and WK soils incubated for two weeks and six months.
195 Halkett soil - two weeks Halkett soil - six months
Cd Cd BCd BCd HM HM BHM BHM MM MM BMM BMM
Treatment LM LM BLM BLM B B C C
0 204060801000 20406080100
Summit soil - two weeks Summit soil - six months
Cd Cd BCd BCd HM HM BHM BHM MM MM BMM BMM
Treatment LM LM BLM BLM B B C C
0 204060801000 20406080100
Wakanui soil - two weeks Wakanui soil - six months
Cd Cd BCd BCd
HM HM
BHM BHM MM MM
BMM BMM
Treatment LM LM
BLM BLM
B B
C C
0204060801000 20406080100
Mean % of Cu species distribution Mean % of Cu species distribution FA bound Cu Cu2+ CuSO4 Remaining inorganic species Figure A13 Mean % of Cu species present in soil solution as colloidal fulvic acid bound (FA), simple 2+ + + + 2 ion (Cu ), CuSO4 and remaining inorganic species (CuOH , Cu(OH)2, CdCl , CuHCO3 , Cu(CO3)2 -) for HK, SM and WK soils incubated for two weeks and six months.
196 Halkett soil - two weeks Halkett soil - six months
Cd Cd BCd BCd HM HM BHM BHM MM MM BMM BMM Treatment Treatment LM LM BLM BLM B B
C C
0 204060801000 20406080100
Summit soil - two weeks Summit soil - six months
Cd Cd
BCd BCd
HM HM
BHM BHM
MM MM
BMM BMM
Treatment Treatment LM LM
BLM BLM
B B
C C
0 204060801000 20406080100
Wakanui soil - two weeks Wakanui soil - six months
Cd Cd
BCd BCd
HM HM
BHM BHM
MM MM
BMM BMM
Treatment LM LM
BLM BLM
B B
C C
0 20 40 60 80 100 0 20406080100
Mean % of Cu distribution in soils Mean % of Cu distribution in soil
Particulate HA Particulate FA Fe-oxides Al oxides Clay Aqueous phase
Figure A14 Mean percentage of Cu distribution on organic matter (particulate HA and FA), Fe and Al oxides, clay, and in the aqueous phase for HK, SM and WK soils incubated for two weeks and six months
197 Halkett soil - two weeks Halkett soil - six months
Cd Cd BCd BCd HM HM BHM BHM MM MM BMM BMM Treatment LM LM BLM BLM B B C C
0 204060801000 20406080100
Summit soil - two weeks Summit soil - six months
Cd Cd
BCd BCd
HM HM
BHM BHM
MM MM
BMM BMM
Treatment LM LM
BLM BLM
B B
C C
0204060801000 20406080100
Wakanui soil - two weeks Wakanui soil - six months
Cd Cd
BCd BCd
HM HM
BHM BHM
MM MM
BMM BMM
Treatment Treatment LM LM
BLM BLM
B B
C C
0 204060801000 20406080100
Mean % of Ni species distribution Mean % of Ni species distribution
FA bound Ni Ni2+ NiSO4 Remaining inorganic species Figure A15 Mean % of Ni species present in soil solution as colloidal fulvic acid bound (FA), simple 2+ + ion (Ni ), NiSO4 and remaining inorganic species (NiOH, Ni(OH)2, NiCl, NiHCO3 , NiCO3) for HK, SM and WK soils incubated for two weeks and six months.
198 Halkett soil - two weeks Halkett soil - six months
Cd Cd BCd BCd HM HM BHM BHM MM MM BMM BMM
Treatment Treatment LM LM BLM BLM B B C C
0 204060801000 20406080100 Summit soil - two weeks Summit soil - six months
Cd Cd BCd BCd HM HM BHM BHM MM MM BMM BMM
Treatment LM LM BLM BLM B B C C
0 20406080100 0 20406080100 Wakanui soil - two weeks Wakanui soil - six months
Cd Cd BCd BCd HM HM BHM BHM MM MM BMM BMM
Treatment Treatment LM LM BLM BLM B B C C
0204060801000 20406080100 Mean % of Ni distribution in soil Mean % of Ni distribution in soil
Particulate HA Particulate FA Fe-oxides Al-oxides Clay Aqueous phase
Figure A16 Mean percentage of modeled Ni distribution on the organic matter (particulate HA and FA), Fe and Al oxides, clay, and in the aqueous phases for HK, SM and WK soils incubated for two weeks and six months.
199 Halkett soil - two weeks Halkett soil - six months
Cd Cd BCd BCd HM HM BHM BHM MM MM BMM BMM LM LM Treatment Treatment BLM BLM B B C C
0 204060801000 20406080100
Summit soil - two weeks Summit soil - six months
Cd Cd BCd BCd HM HM BHM BHM MM MM BMM BMM
Treatment LM LM BLM BLM B B C C
020406080100 0 20406080100
Wakanui - two weeks Wakanui soil - six months
Cd Cd BCd BCd HM HM BHM BHM MM MM BMM BMM Treatment LM LM BLM BLM B B C C
0 204060801000 20406080100 Mean % of Zn species distribution Mean % of Zn species distribution
FA bound Zn Zn2+ ZNSO4 Remaining inorganic species Figure A17 Mean % of Zn species present in soil solution as colloidal fulvic acid bound (FA), simple 2+ + + + ion (Zn ), ZnSO4 and remaining inorganic species (ZnOH , Zn(OH)2, ZnCl , ZnHCO3 , ZnCO3) for HK, SM and WK soils incubated for two weeks and six months.
200 Halkett soil - two weeks Halkett soil - six months
Cd Cd BCd BCd HM HM BHM BHM MM MM BMM BMM
Treatment LM LM BLM BLM B B C C
0 20406080100 0 20406080100
Summit soil - two weeks Summit soil - six months
Cd Cd BCd BCd HM HM BHM BHM MM MM BMM BMM
Treatment LM LM BLM BLM B B C C
0 20406080100 0 20406080100 Wakanui soil - two weeks Wakanui soil - six months
Cd Cd BCd BCd HM HM BHM BHM MM MM BMM BMM
Treatment LM LM BLM BLM B B C C
0 20406080100 0 20406080100
Mean % of Zn dsitribution in soil Mean % of Zn distribution in soil
Particulate HA Particulate FA Fe-oxides Al-oxides Clay Aqueous phase
Figure A18 Mean percentage of Zn distribution on organic matter (particulate HA and FA), Fe and Al oxides, clay, and in the aqueous phase for HK, SM and WK soils incubated for two weeks and six months.
201 Appendix B. Plant results for Chapter 4
Germination rates after 5 days Biomass of wheat shoots 1.2 20 Halkett soil 1.0
15 0.8
0.6 10
Dry weight(g)Dry 0.4
5 0.2 Seedling emergence (from 22 seeds) 0 0.0 C B BLM LM BMM MM BHM HM BCd Cd C B BLM LM BMM MM BHM HM BCd Cd
1.2 Summit soil Summit soil 20 1.0
15 0.8
0.6 10
Dry weightDry (g) 0.4
5 0.2 Seedlingemergence (from 22 seeds) 0 0.0 C B BLM LM BMM MM BHM HM BCd Cd CBBLMLMBMMMMBHMHMBCdCd
1.2 20 Wakanui soil Wakanui soil 1.0
15 0.8
0.6 10
Dry weight(g)Dry 0.4
5 0.2 Seedling emergence (from 22 seeds) 0 0.0 C B BLM LM BMM MM BHM HM BCd Cd C B BLM LM BMM MM BHM HM BCd Cd
Treatment Treatment Two weeks Six months
Figure A19 Mean germination rates and dry weights of wheat seedlings grown in biosolids amended and unamended metal treated soils incubated for two weeks and 24 weeks at field capacity. Bars denote the SEM, n = 89 for both trials.
202 Cd concentration in wheat shoots A B
Halkett soil Summit soil CL CL ) ) -1 -1
40 40
20 20 Cd concentration (mg kg Cd concentration (mg kg (mg concentration Cd 10 10
0 0 C B BLM LM BMM MM BHM HM BCd Cd C B BLM LM BMM MM BHM HM BCd Cd C
Wakanui soil Two weeks CL Six months ) -1 Abbreviation key C = Control B = Biosolids only 40 BLM = Biosolids + low metals 20 LM = Low metals BMM = Biosolids + medium metals MM = Medium metals BHM = Biosolids + high metals Cd concentration (mg kg 10 HM = High metals BCd = Biosolids + Cd only Cd = Cd only
0 C B BLM LM BMM MM BHM HM BCd Cd Treatment
Figure A20 Mean Cd concentrations for in wheat shoot tissue (mg kg-1 dry weight) grown in biosolids amended and unamended metal spiked soils incubated for two weeks and 24 weeks. The critical level (CL) for Cd as depicted by the red line is 43 mg kg-1 (Macnicol and Beckett, 1985). Bars denote the SEM, n = 89 for both trials.
203 Cu concentration in wheat shoots A B 100 100 Halkett soil Summit soil ) 80 ) 80 -1 -1
60 60
40 40 Cu concentration (mgkg CL (mg kg Cu concentration CL 20 20
0 0 C B BLM LM BMM MM BHM HM BCd Cd C B BLM LM BMM MM BHM HM BCd Cd
C 100 Wakanui soil Two weeks Six Months
) 80 -1
60 Abbreviation key C = Control B = Biosolids only BLM = Biosolids + low metals 40 LM = Low metals BMM = Biosolids + medium metals MM = Medium metals
Cu concentration (mg kg CL 20 BHM = Biosolids + high metals HM = High metals BCd = Biosolids + Cd only
0 Cd = Cd only C B BLM LM BMM MM BHM HM BCd Cd Treatment
Figure A21 Mean Cu concentrations in wheat shoots (mg kg-1 dry weight) grown in biosolids amended and unamended metal spiked soils incubated for 2 weeks and 24 weeks. The critical levels (CL) for Cu as depicted by the red line, is 17 mg kg-1 (Macnicol and Beckett, 1985). Bars denote the SEM, n = 89 for both trials.
204 Ni concentration in wheat shoots A B 160 Halkett soil 600 Summit soil 140
) 500 ) 120 -1 -1
400 100
80 300
60 CL 200 Ni concentration (mg kg Ni concentration (mg kg concentration Ni 40
100 CL 20
0 0 C B BLM LM BMM MM BHM HM BCd Cd C B BLM LM BMM MM BHM HM BCd Cd
C 160 Wakanui soil Two weeks 140 Six months
) 120 -1
100 Abbreviation key C = Control B = Biosolids only 80 BLM = Biosolids + low metals LM = Low metals 60 BMM = Biosolids + medium metals CL MM = Medium metals
Ni concentration (mg kg 40 BHM = Biosolids + high metals HM = High metals
20 BCd = Biosolids + Cd only Cd = Cd only
0 C B BLM LM BMM MM BHM HM BCd Cd Treatment Figure A22 Mean Ni concentrations in wheat shoots (mg kg-1 dry weight) grown in biosolids amended and unamended metal spiked soils incubated for 2 weeks and 24 weeks. The critical levels (CL) for Ni as depicted by the red line, is 46 mg kg-1 (Macnicol and Beckett, 1985). Bars denote the SEM, n = 89 for both trials.
205 Zn concentration in wheat shoots A B 1600 400 Halkett soil Summit soil 1400 ) ) 1200 -1
-1 300
1000 CL
800 200
600
400 100 Zn concentration (mg kg Zn concentration (mg kg CL 200
0 0 C B BLM LM BMM MM BHM HM BCd Cd C B BLM LM BMM MM BHM HM BCd Cd
C 400 Wakanui soil Two weeks Six months )
-1 300 Abbreviation key CL C = Control B = Biosolids only 200 BLM = Biosolids + low metals LM = Low metals BMM = Biosolids + medium metals MM = Medium metals BHM = Biosolids + high metals 100 Zn concentration (mg kg HM = High metals BCd = Biosolids + Cd only Cd = Cd only
0 CBBLMLMBMMMMBHMHMBCdCd Treatment Figure A23 Mean Zn concentrations in wheat shoot tissue (mg kg-1 dry weight) grown in biosolids amended and unamended metal spiked soils incubated for 2 weeks and 24 weeks. The critical levels (CL) for Zn as depicted by the red line, is 224 mg kg-1 (Macnicol and Beckett, 1985). Bars denote the SEM, n = 89 for both trials.
206 Appendix C. Regression and ANOVA results for DOC, salinity and shoot Cd concentrations for Chapter 4
Regression analysis for the relationship determined between DOC and shoot Cd concentrations
R = 0.27966490 Rsqr = 0.07821246 Adj Rsqr = 0.06465676
Standard Error of Estimate = 0.3952
Coefficient Std. Error t P y0 -0.6595 0.4426 -1.4900 0.1409 a 0.4624 0.1925 2.4020 0.0190
Analysis of Variance: DF SS MS F P Regression 1 0.9011 0.9011 5.7697 0.0190 Residual 68 10.6204 0.1562 Total 69 11.5215 0.1670
Regression analysis for the relationship determined between soil solution salinity and shoot Cd concentrations
R = 0.44867972 Rsqr = 0.20131349 Adj Rsqr = 0.18956810
Standard Error of Estimate = 0.3679
Coefficient Std. Error t P y0 -1.0031 0.3412 -2.9401 0.0045 a 0.6044 0.1460 4.1400 <0.0001
Analysis of Variance: DF SS MS F P Regression 1 2.3194 2.3194 17.1398 <0.0001 Residual 68 9.2021 0.1353 Total 69 11.5215 0.1670
224